DNA signatures preserved in the official 1978 sample collection of the Shroud of Turin

The Shroud of Turin is a linen burial cloth, approximately 4.4 m long and 1.1 m wide, which shows the double bodily image, front and back, of a man who has apparently suffered physical trauma through crucifixion with signs that are interpreted and believed to be compatible with those described in the New Testament^1,2. It has been venerated for centuries, especially by members of the Catholic Church, and popular belief has this artifact as the cloth used to wrap the lifeless body of the biblical Jesus of Nazareth, who is acknowledged as the founder of the Christian movement that started in the Judean region in the mid-first century Common Era (CE).

The true nature of the Shroud has been highly debated among historians, theologians, and scientists. Several hypotheses regarding its origin have been proposed, but as of today, no one has yet been able to establish to any degree of certainty when and where this textile originated from. The only historical fact that has been recorded is that the Shroud has been in Europe since the mid-fourteenth century, when it first appeared in Lirey, France, between 1353 and 1356. Starting in 1453, it was kept in Chambéry by the Dukes of Savoy until 1578, followed by several moves between France and the northern part of Italy during the sixteenth century, with the final resting place in Turin. Kept in a reliquary from 1694 to 1993, the Shroud was held permanently in the chapel adjacent to the Turin Cathedral, except for a brief move to Genoa in 1706 and to Montevergine, in the province of Avellino, during the Second World War. After a fire in the chapel in 1997, the Shroud was moved to the left chapel in the transept of the Turin Cathedral (for details, see^3,4,5).

Considering available information and material, it has not yet been possible to either confirm or refute that the Shroud existed before its appearance in France in the mid-fourteenth century. However, a commonly held belief is that its journey through the years began in Jerusalem, in a tomb in the Holy Land, approximately 2,000 years ago, in the year 31 or 33 CE (⁴ and references therein). Tradition narrates that after it was concealed for years, the Shroud would have been kept in what is modern-day Turkey, first in Mesopotamia then in Edessa, followed by a transfer to Şanliurfa, possibly around 200 CE. Subsequent moves would have seen the Shroud in Anatolia, Constantinople (modern-day Istanbul), in 944 CE⁵. A burial cloth, which some historians believe to be the Shroud, was owned by the Byzantine emperors but disappeared during the sacking of Constantinople in 1204 CE⁶. After this event, presumably the Shroud would have been taken by crusaders and transferred to Athens, Greece, where it would have remained until 1225. Evidence for the transfer of the Shroud from Constantinople to Athens is provided by chronicles of the Fourth Crusade^7,8,9,10.

Iconographic representations on Byzantine coins, the so-called gold solidi used during the Eastern Roman Empire, portray the face of the Man of the Shroud during the time of Justinian II, the Byzantine emperor who lived between 669 and 711 CE. Numismatic experts believe that the engraver must have seen the Shroud before creating the image shown on this coin because of some distinctive facial traits in common with the sacred relic. Among the various historical proofs, it is worth mentioning the votive medallion of Lirey, found in France, which shows the frontal and dorsal image of the Man of the Shroud, probably dated between 1356 and 1389. In 1988, the age of the linen cloth of the Shroud of Turin was assessed by accelerator mass spectrometry, and results of radiocarbon measurements from separate independent laboratories yielded a calendar age range of 1260–1390 CE, thus suggesting a recent Medieval origin¹¹. However, the raw data of the radiocarbon measurements were not initially released and were made accessible through the British Museum only in 2017. Statistical analysis of the original raw data strongly suggests an absence of homogeneity between the datasets made by the three laboratories, which may have undermined their reliability¹², thus requiring additional testing to ascertain its age.

The possible existence of the Shroud prior to the first documented information places the long journey of this artifact into a Middle or Near East geographical context, with a potential historical age preceding the Sacking of Constantinople in 1204. This would have then been followed by a Western European relocation and subsequent reappearance at Lirey in 1353 or 1356. Our current study attempts to differentiate between the two referenced periods separated by the so-called “missing years”¹³, considering the two geographical areas characterized by distinctive genomic data associated with human lineages, plant species, or/and animal breeds. Furthermore, we divide the chronological history of this artifact into a recognizable “pre-1204 CE”, which is plausibly associated with any geographical region within the ancient Near East, and a plausible “post-1353 CE”, whose location is likely restricted to Western Europe. These two hypothetical temporal and spatial differentiations may be reflected in the variation of DNA data obtained from the Shroud.

Genomic DNA was previously isolated and analyzed from organic residues recovered in 1978 from different parts of the body image on the back of the Shroud, and from threads taken from the edges and used in 1988 for radiocarbon dating of this sacred artifact¹⁴. Analysis of the DNA extracted from organic particles and textile strands sampled from the Shroud revealed chloroplast and mitochondrial genome sequences representing multiple plant species, as well as various human biogeographical origins. In particular, these analyses revealed the presence of at least 19 plant species common in the Mediterranean Basin, in addition to plants whose primary center of origin is Asia (China), the Middle East, and the Americas, some of which reflects their introduction to the “Old World” after the twelfth century. In addition, the analyses revealed mtDNA sequences from at least 14 people of different geographic ancestry, based on the Eurasian haplogroups identified which are common to Western Europe and North-Eastern Africa, as well as from the Middle East, the Arabian Peninsula, the Caucasus region, and also rare haplotypes from the Indian sub-continent¹⁴. A quantitative assessment of the overall amplicon sequencing data revealed that over 55.6% of the human DNA corresponds to lineages from the Near East, while Western European lineages account for less than 5.6%. The presence of 38.7% of the overall human genomic data from Indian lineages is unexpected and is potentially linked to historical interactions associated with importing linen or yarn from regions near the Indus Valley, referred to as “Hindoyin” according to rabbinic texts¹⁵.

The present study aimed to isolate and characterize genomic DNA sequences from Turin Shroud samples using a different molecular methodology (metagenomic analysis using PCR-free NGS protocols, whereas the previous study relied on PCR amplification of mtDNA fragments) and an original sample consisting of linen filaments collected in 1978 by the late Prof. Pierluigi Baima Bollone from various sections of the Turin Shroud^16,17. After DNA extraction in clean rooms, shotgun genomic libraries were sequenced using an NGS strategy with the main goal to determine the quantity and quality of human DNA and any residual DNA from microbial, fungal, plant, and animal species. The genetic results obtained were then correlated with historical information, the corresponding geographical areas of origin, and the modern distribution of plant species and human geographic groups with the aim of acquiring new data and providing novel clues as to the history of the Shroud of Turin.

Human DNA sequences identified on the Turin shroud

Before testing a total of twelve samples collected from the Turin Shroud (TS) (Table S1 and Figure S1), we verified the possibility of isolating and sequencing human DNA fragments from a linen cloth that had been purposely contaminated by touch. We obtained a low-coverage genome (~ 1.8 ×) from the human control (lab operator; Dataset S1A).

We extracted DNA, built genomic libraries and obtained genomic reads from seven TS samples (Fig. 1 and Dataset S1B).

The percentage of trimmed reads that mapped to the human reference genome exceeded 10% in three cases, resulting in an average depth higher than 0.5 × (higher than 1 × for AB1 and B12). The average length of the reads (less than 100 base pairs) is consistent with the characteristics of degraded, possibly ancient DNA. However, both the damage patterns and the contamination tests show clear signatures of modern DNA contamination (Dataset S1B), as expected given the large number of people who have come into contact with the TS. To further characterize the possible presence of ancient DNA, we extracted, from each sample, reads showing patterns of deamination at their terminal bases. We applied the maximum-likelihood probabilistic model implemented in PMDtools¹⁸ to assign post-mortem damage (PMD) scores to each read. The distribution of PMD scores is very similar among the seven samples, with most reads (> 98.3%) showing a score lower than 1 (Figure S2). Only a small fraction showed PMDs greater than 3 (0.21–0.53%), which could be considered the minimum threshold for reliable ancient reads. For a threshold of PMDs > 5, between 0.05 and 0.19% of the total sequences were retained, comparable to values found in modern and historical individuals¹⁸. An example of the damage patterns at different PMD scores is reported in the supplementary information for sample AB1 (Figure S3).

The genetic sex from these reads was determined to be likely that of a biological male (Dataset S1B), and a predominant mitochondrial DNA (mtDNA) haplogroup was determined with high accuracy (Q > 0.95; Dataset S2A) in samples AB1, B12, and C10-gl. This haplogroup, corresponding to K1a1b1a, is a typical Ashkenazi subclade of K1a1b1 with a likely Western European origin¹⁹, but was also found among Sephardic Jews. The same haplogroup, and the same haplotype, were identified by analyzing the mitochondrial genome of the late Prof. Baima Bollone (the collector) (Dataset S2B). Therefore, we performed kinship analyses, which revealed a close relationship (same individual or first-degree) between the three TS DNAs and the genome of the collector, who has Jewish ancestry (Table S2). As a negative control, we confirmed the complete unrelatedness between the TS genomic data (and the collector) and the modern linen specimen, artificially contaminated by the operator.

Other haplogroups were also identified in the three samples (Figs. 1 and 2), namely H2a2, H33 and H1b. The first two were also reported by Barcaccia et al.¹⁴. H2a2 is expected since it is the lineage of the mitochondrial reference sequence rCRS (NC_012920) and therefore can be identified in any sample without distinctive haplogroup markers. H1b is commonly found in Europe, and to a lesser degree also in Central Asia²⁰. H33 is a rare haplogroup found today also in Near Eastern groups, especially among the Druze, an Arabic-speaking ethno-religious minority currently present in the Holy Land, Jordan, Lebanon, and Syria^21,22.

Many variants were identified as heteroplasmic (Fig. 2 and Dataset S2A). Some of these are located at known mutational hotspots, some occur at haplogroup-defining positions, and some may represent private mutations. Several of the previously reported mitochondrial variants were identified among these heteroplasmies. However, no additional haplogroups could be assigned with high accuracy. This is due to the high-quality filtering used for variant calling (i.e., 176 heteroplasmic variants are excluded from the final haplotypes) and the ability to analyze entire mitogenomes, which greatly increases the accuracy of classification, up to more than 70%²³. Our results indicate that the human DNA found on the TS specimens mostly derives from the main collector and shows variable levels of additional contamination.

Proteomics analyses

We used liquid chromatography–tandem mass spectrometry to analyze the peptide profiles of three TS samples (namely, C/D12, B12, and R58). Analyses run with ProgenesisQI for Proteomics revealed, after excluding four proteins identified in the negatives—keratin type II cytoskeletal 1, triosephosphate isomerase, actin, and trypsin (deemed normal contaminants or deliberately added to the sample)—and other common contaminants, such as myosin and keratins 9 and 10, the only unique proteins identified in the C/D12 and R58 samples were proteinase K, with 16 unique peptides identified overall, and collagen alpha-1(I) chain, with two unique peptides identified in the R58 sample (Dataset S3). Analyses run with FragPipe were consistent with these data and additionally identified collagen alpha-1(I) chain in both samples C/D12 (1 unique peptide) and B12 (1 unique peptide), and fibroin heavy chain from Bombyx mori (5 unique peptides) only in R58 sample. The presence of silk-associated proteins is consistent with either contamination by silk or historical contact with silk materials. However, it does not provide evidence that the shroud itself is composed of silk, given that the textile is known to be made of linen. The presence of proteinase K in the samples was expected because the TS samples had been treated with proteinase K during previous attempts at DNA extraction. Since the peptides of the collagen alpha-1(I) chain showed no modification, they likely originated from skin contamination rather than from ancient peptides. This is further supported by the high presence of keratins, which are much higher in the samples than in the negative controls, indicating poor handling procedures rather than issues associated with laboratory analysis. Overall, it can be concluded that no ancient peptides were identified in the samples, only keratins and other proteins associated with skin handling.

Radiocarbon dating

Two linen threads collected from the reliquary were quantitatively sufficient for 14C radiocarbon dating (R58-1 and R58-2, respectively), which was outsourced to the Curt-Engelhorn-Zentrum Archäometrie (CEZA) in Mannheim, Germany (Fig. 3). The obtained ages (Dataset S1C) are consistent with two historical events. The time range of sample R58-1 (1451–1622 CE) overlaps with the fire that broke out in 1532 CE in the Chapel of Saint-Chapelle in Chambéry, where the Turin Shroud was being kept inside a silver reliquary. The intense heat caused by contact with molten silver burned through the folded cloth. The Shroud was saved from destruction, and in 1534 CE, Poor Clare nuns repaired the burn holes by sewing linen (or silk) patches over the damaged areas²⁴. Further conservation mending work was carried out in 1694 CE, likely under the direction of Sebastian Valfrè, to stabilize the Shroud²⁵. This work overlaps with the time range of sample R58-2 (1642–1800 CE).

Metagenomics analyses

Due to the low number of reads from the Turin Shroud samples that mapped to the human reference genome, we conducted a comprehensive metagenomic analysis. We utilized the Kraken program for the initial screening. The distribution of reads across kingdoms varied significantly (Table S3, Dataset S5), with unassigned reads ranging from 17% (AB1) to 72% (R58). Reads assigned to viruses and archaea were minimal, while bacterial species consistently accounted for 10–31% of the reads. The percentages of eukaryotic reads differed significantly and showed an inverse relationship with unassigned reads; samples with fewer unassigned reads had a higher proportion of eukaryotic reads and vice versa. Regarding the eukaryotic reads, the vast majority were assigned to the genus Homo: 93% or more in AB1, C/D12, C10_gl, and B12; 86% in C3_ha; and 77% in R58 (Table S3, Dataset S5). A notable exception was sample C12_fe, containing only 0.4% eukaryotic reads, of which 26% were assigned to Homo. As in the other samples, most of the remaining reads were assigned to Viridiplantae (e.g., 61% in the case of C12_fe).

The data derived from the Kraken classification enabled identification of Bacteria, Archaea, and Fungi, but were insufficient for Metazoa and Viridiplantae analyses due to limited sequence databases. Therefore, we adopted an alternative approach based on aligning the sequences against the NCBI Nucleotide (nt) database. Because running BLAST on all individual reads was not feasible due to their high number and short length, we performed a metagenomic assembly of the reads and subsequently aligned the resulting contigs to the database using BLAST. Challenges due to species diversity and DNA degradation resulted in low assembly metrics and short contigs (with an average length of 210–250 nt, Table S4). The results of the BLASTn analysis were categorized as those from Kraken (Table S5), showing comparable trends with generally fewer unassigned contigs due to the larger NCBI Nucleotide database used. Samples R58 and C12_fe were exceptions with higher unassigned contig shares (83.8% and 95.1%, respectively). Both archaea and viruses were present only in trace amounts. Compared to the Kraken analysis, eukaryotic contig proportions increased in all samples, likely because of assembly biases that favored the abundant Homo sapiens genome. A significant number of bacterial contigs could not be classified at the species or phylum levels. Only a few archaea taxa were detected (a maximum of 28 in B12 and none in R58).

The fungi kingdom was used for comparison in the analyses. Identified orders showed strong similarity across samples, although less represented orders exhibited some variability. Kraken was deemed more reliable for species quantification, while BLAST performed better for species identification despite potential false negatives.

The Kraken results were further validated through an additional analysis with MetaPhlAn, a tool specifically designed to profile microbial communities from metagenomic shotgun sequencing²⁶. Unlike Kraken, which uses k-mer analysis across entire genomes, MetaPhlAn aligns unpaired reads to a comprehensive microbial gene database for precise identification. Consequently, MetaPhlAn identified fewer reads, with unclassified readings from 44 (AB1) to 99% (R58 and C12_fe). The detected taxa mostly aligned with the findings derived from Kraken analysis, identifying minimal archaea and the same bacterial classes. New species constituted only 2–4% of the reads in samples R58 and C12_fe, and were attributed to Geobacillus jurassicus, a thermophilic bacterium²⁷ (Dataset S6). The concordance of the outputs obtained with two different methodologies (and reference databases) reinforces our results.

Bacterial and Fungal taxa identified on the Turin Shroud

Overall sequencing data for bacterial, archaeal, and fungal biota, including the relative abundance of the phyla, classes, and orders, respectively, are summarized in Fig. 4, with further details provided in the supplementary information (Datasets S4 and S5).

Considering current knowledge of microbial ecology, the analysis of bacterial taxa associated with the TS samples revealed a notable presence of species commonly found on the human skin, particularly in sample AB1 and to a lesser extent in sample C3_ha. These include Cutibacterium acnes (formerly Propionibacterium acnes), Staphylococcus epidermidis, Cutibacterium granulosum, Corynebacterium tuberculostearicum, Staphylococcus capitis, and Staphylococcus hominis²⁸. Their presence indicates potential human skin contact with the analyzed samples. However, analyses of damage patterns of reads aligned to the identified contigs with PyDamage²⁹ revealed no detectable amounts of deamination.

Within the archaeal domain, the TS samples exhibited strong consistency in community composition, with most taxa recurring among the top entries. These predominantly belonged to the salt-tolerant class Halobacteria, which is adapted to high salinity environments. These findings suggest a pronounced overrepresentation of osmotically resilient species, which are typically associated with saline and drought-stressed conditions.

The fungal community displayed a significant abundance of Debaryomyces hansenii, a yeast species that thrives in dry and osmotically stressful environments³⁰, similar to the environments where the Turin Shroud was kept in the past. Saprophytic molds, such as Aspergillus, were also identified, along with fungi associated with human skin and scalp, particularly Malassezia restricta.

Taxonomic identification of a diverse array of bacterial, archaeal, and fungal taxa provided insight into the complex microbiome associated with the TS, shedding light on possible ecological and historical contexts. Bacterial profiles featured prominently skin-associated taxa, which indicate direct human contact, both in ancient and modern times. The high concentration of these bacteria in sample AB1, which is from a peripheral area likely exposed to environmental contamination, suggests frequent manipulation. Sample C/D12, taken from the hand region of the Shroud’s image, further supports evidence of human contact. The detection of Paraburkholderia species, known for their ecological adaptability and occasional pathogenicity³¹, may reflect broader environmental exposures over time.

Archaeal analysis revealed a consistent dominance of Halobacteria, which suggests an adaptation to saline environments. This could be related to ancient flax retting practices in brackish or salty water³², which were more common than dew retting in antiquity³³. The dominance of Debaryomyces hansenii further supports salt tolerance and possible marine connections because the yeast is also found in seawater³⁰. The presence of Aspergillus points to possible degradation related to fluctuating storage conditions. The widespread presence of Malassezia restricta, a primary colonizer of the human skin³⁴, across all samples is a clear demonstration of ongoing human interaction with the Shroud.

Overall, the microbiological diversity of the Shroud samples offers valuable insight into its history. The recurrent presence of skin-associated bacteria and fungi implies direct contact with human beings, possibly during devotional religious ceremonies and cloth conservation treatments, in addition to the official sampling operations. These findings underscore the Shroud’s dual significance as a religious relic and a historical artifact, reflecting both devotional and environmental influences throughout its existence.

Plant and animal taxa identified on the Turin Shroud

A metagenomic assembly of the reads into contigs (see Material and Methods) and a nucleotide BLAST search enabled us to extend the analysis to other kingdoms (Fig. 5, and Dataset S7 and Dataset S8).

In addition to Homo sapiens, our analysis of animal taxa revealed notable findings, particularly regarding the red coral (Corallium rubrum; 33.3%), a species endemic to the Mediterranean Sea (Fig. 5B-C, and Dataset S7 and Dataset S8). However, reads mapped to the reference genomes of the identified animal and plant taxa represented by more than 100 contigs exhibited no detectable signatures of terminal cytosine deamination, including reads with PMDS > 5 (Figures S4 and S5). Comparable results were obtained through independent analysis of the reads mapped to the contigs using PyDamage²⁹.

We focused on the taxa whose identification was based on hits with at least 100 assigned contigs to avoid false positives, and found domestic cats and dogs to be prevalent (21.7% and 22.1%, respectively), although DNA traces from several farm species, including chickens (Gallus gallus domesticus; 9.5%) and ruminants from the Bovidae family (cattle, goats, and sheep; 2,6% overall) were detected. Other livestock and wild animals such as pigs, horses, deer, and rabbits were also found. Weak traces of brown rats and mice suggest possible environmental contamination. Among the less abundant taxa identified (around 1% of assigned contigs), notable are some fish species including the gray mullet (Chelon labrosus), Atlantic cod (Gadus morhua), and ray-finned fishes (Leuciscinae spp.). Marine ostracods, particularly the small seafloor-dwelling crustacean Notodromas monachal, which is common in zooplankton communities, were also documented. Insects such as flies (Diptera), aphids (Hemiptera), and arachnids like dust and skin mites and ticks, including Demodex folliculorum and Haplochthonius simplex, were identified as well.

Plant taxa showed greater diversity (Fig. 5D, E, and Dataset S7 and Dataset S8), with prominent species including carrot (Daucus carota; 30.9%) and bread wheat (Triticum aestivum; 11.6%), represented by more than 100 assigned contigs. Other cereals found encompassed durum wheat (Triticum durum), einkorn wheat (Triticum monococcum), maize (Zea mays) and rye (Secale cereale). Horticultural crops included peppers, tomatoes, and potatoes (Solanaceae), as well as melons or cucumbers (Cucurbitaceae). The Fabaceae family was also represented by a strong presence of peanuts (Arachis spp.), particularly in sample C/D12. Among forage species and their wild relatives, weak traces of perennial ryegrass, bluegrass, fescue, oats, and clovers were observed. The fruit species identified with greater confidence included banana, almond, walnut, and sweet orange. Fainter DNA signals were detected for fig, pistachio, apple, pear, hazelnut, maple, and grapevine (Vitis spp.). The only non-cultivated plant with more than 100 assigned contigs was stinging nettle, but significant amounts of spruce trees and wild cardoons were also found (see Dataset S7 and Dataset S8).

Comparison against a database of known contaminated sequences revealed that nearly all contaminated RefSeq or GenBank entries detected among the BLAST hits were of prokaryotic origin, with only a few exceptions (Dataset S9). Specifically, Sporisorium reilianum (Fungi, Ustilaginaceae), Botryotinia fuckeliana (Fungi, Sclerotiniaceae), and Cyprinus carpio (Metazoa, Leuciscidae) were absent from the final contig assignment list. This indicates that the contigs initially matching these taxa also exhibited significant similarity to other members of the same family or order that were retained in the final assignment. The identification of sequence regions shared among multiple closely related taxa therefore suggests that these matches do not result from contamination originating from phylogenetically distant species. By contrast, Solanum lycopersicum, Solanum pennellii (Viridiplantae, Solanaceae), and Ipomoea triloba (Viridiplantae, Convolvulaceae) were also represented in the final assignment list, although the latter two were each detected only once. Nevertheless, most contigs matching these species were likewise shared with closely related taxa. Consequently, we conclude that any potential contamination associated with these records is unlikely to have substantially influenced the taxonomic assignments.

The taxonomic identification of the plant and animal species on the Shroud reveals a complex biodiversity, albeit constrained by limitations in available reference databases. Reliance on a relatively small dataset (126 plant species and limited animal genomes) and the degraded condition of the DNA available made the taxonomic classification complicated. Nevertheless, several Viridiplantae taxa were confidently identified, particularly agricultural species such as carrots and cereals. Grouping of taxa at the genus level (e.g., Triticum, Aegilops, Hordeum, and Secale spp.) reflects a parsimonious approach due to the overlapping of genetic signals. However, the strong signal for Daucus carota indicates that higher specificity could be obtained in some cases.

Among cultivated species, we also identified melon or cucumber, and various Solanaceae (tomato, pepper and potato). These reflect typical agricultural practices in the Mediterranean and other European territories. Fruit species such as banana (Musa acuminata) and orange (Citrus sinensis) further highlight the diversity of cultivated plants possibly introduced through trade and indicate interconnected regional economies.

Non-cultivated species such as cardoon (Cynara cardunculus), maple (Acer spp.), spruce (Picea spp.), pine (Pinus spp.), nettle (Urtica dioica), and walnut (Juglans spp.) reflect local ecological diversity. Notably, cardoon and nettle are indigenous to Europe and the Mediterranean area. Wild cardoon remains widespread in southern Europe and North Africa, and it was domesticated in the western Mediterranean^35,36. Stinging nettle, native to Europe and widely distributed across Asia and North Africa, is valued both as a vegetable and in traditional Indian medicine³⁷. The uncertain origin of melon is also noted, with hypotheses pointing to ancient Persia or Africa³⁸.

Although none of the identified plant taxa can be uniquely associated with the Levant and the Middle East, several genera originated in the Fertile Crescent and are now widespread across the Mediterranean Basin. For example, almond trees, pistachios, and figs are native to and have been domesticated in the Near or Middle Eastern areas^39,40,41, following the introduction by Romans to Southern Europe and Northern Africa during the first centuries CE.

The presence of exotic species introduced to Europe from Central and Eastern Asia, and North and South America, was also noted. Maples, for instance, are thought to have originated in Asia, with genetic studies suggesting that the genus Acer likely emerged in East Asia, particularly in China. However, fossil evidence indicates that some species may have evolved independently in North America⁴². Bananas, whose secondary center of diversification is Africa, were introduced to Europe by Portuguese sailors in the early sixteenth century, but they gained popularity only by the eighteenth century. This shift was driven by improved agricultural practices and trade networks, including plantation systems in Central America (e.g., the Caribbean), which helped establish bananas as a staple fruit in Europe⁴³.

Several plant taxa identified here, such as Picea abies, Juglans regia, Pinus spp. Cucumis spp., Cichorium spp., Cynara spp., and Vitis spp., were also found in previous palynological and/or molecular studies^14,44,45,46. These findings support a view of the Shroud as being associated with a blend of local vegetation and historical agricultural practices.

Animal taxa presented a more complex identification challenge due to the genetic similarities among related species. The presence of Corallium rubrum underscores the region’s historical reliance on marine resources. This endemic Mediterranean red coral has been known since the time of the Roman Empire and held considerable cultural, economic, and symbolic significance in antiquity. It was primarily fashioned into jewelry, amulets, talismans, and decorative objects. Several ancient texts from the Middle Ages attest to its use in medical contexts, as it was believed to possess healing properties and was utilized in remedies for various ailments and disorders. Regarded as a valuable commodity, red coral was transported for centuries along Mediterranean trade routes, facilitating economic and cultural exchanges among populations and regions. The trade of red coral saw further development during the medieval period, particularly in Mediterranean ports^47,48.

Livestock species as well as cats and dogs, reflect long-standing agricultural and domestication practices in local farms and rural settings. The presence of mites and ostracods, including the abundant Notodromas monacha, which are very common in terrestrial and aquatic habitats illustrates the ecological diversity and may reflect environmental conditions during the storage or handling of the Shroud.

Genetic affinities among the Turin Shroud samples

To highlight potential similarities among the TS samples, we performed two different analyses based on their metagenomic profiles (Fig. 6). The artificially contaminated linen fragment was used as an outgroup.

The Principal Component Analysis (PCA) revealed the outlier behavior of AB1, which represents the TS later edge, while samples derived from the areas of face, hands, gluteal region, and feet of the body image of the Man of the Shroud did not show distinct patterns related to sampling type (linen fragment vs. vacuum dust) (Fig. 6A). On the other hand, C12_fe and R58, collected from the Shroud and the Reliquary, respectively, were the samples with the highest number of private taxa in the hierarchical clustering plot (Fig. 6B). It is important to note that the outgroup used as control was confirmed as outlier by both analyses and its genetic differentiation from each of the TS samples was comparable. In addition, although the sum of the first two components accounted for about 72% of the overall genetic variation, most TS samples were closely clustered by PCA, except for AB1. This low discrimination is partially due to the method, which could obscure differences between samples (see Materials and Methods). However, a closer examination of the overall genetic data (Dataset S7) reveals the presence of both different and similar organisms among the pairwise comparisons of the TS samples. These include several domesticated species, such as cereals, vegetables, pets, and farm animals. We believe that this commonality can be explained by two hypotheses rather than solely by the sample history. The first hypothesis is technical: these species are well-described and included in various databases, thus any misalignment during identification is likely to occur with higher frequency. The other hypothesis calls for secondary contamination, which refers to DNA traces resulting not from the direct exposure of the TS samples to those particular species, but rather from human contact with those organisms. Basically, we could hypothesize the presence of a sort of secondary metagenomic fingerprint. This would lead to a collection of species that appear with greater probability than others. However, this hypothesis could not be explored further within our project because it would require a comparison of additional samples (similar to the TS) from different origins.

The presence of recurrent species, while potentially introducing background noise, must be emphasized: numerous other species were identified exclusively in specific samples of the Shroud or with exceptional clarity, such as Daucus carota and Corallium rubrum.

Reappraisal of the overall biodiversity found on the Turin Shroud

The study aimed to isolate and characterize genomic DNA sequences from the available original official samples of the Turin Shroud collected in 1978 by the late Prof. Pierluigi Baima Bollone from various sections of this linen cloth. Its final goal was to identify the biological sources of environmental contaminants—including pollen, plant fibers, fungal structures, animal hairs, blood residues, and human biological fluids—found in linen filaments, a vacuum dust sample, and a small cloth fragment.

Our new findings, alongside previous ones¹⁴, confirm the presence of plant and animal DNA, as well as traces of DNA from various human subjects on the Shroud of Turin. The identification of diverse DNA sources provided insights into potential connections between the temporal and geographic pathways related to hypotheses on the Shroud’s origin. Many identified plant species are native to and widespread in Central Europe and the Mediterranean Basin, extending from the Iberian Peninsula to the Middle East.

The most prominent crop identified in the TS samples is carrot (30.9%). Both historical records and molecular analyses suggest that carrots, characterized by purple and yellow roots, were domesticated in Central and Minor Asia, and then introduced to Western Europe^49,50,51. It is worth mentioning that out of approximately 3800 contigs, about half were confidently mapped to unique sites of the carrot genome, whereas the remaining half mapped to multiple locations being likely associated with repetitive sequences. The majority of the contigs originated from the C10 sample, comprising 1787 contigs (94.4%), whereas only 2 and 1 contigs are from AB1 and B12 samples, respectively. SNPs common to a DNA sequence dataset of carrot accessions⁵² belonging to five distinct populations were employed to calculate their similarity. Based on this analysis, carrot DNA contaminants found on the Shroud were shown to be genetically more similar to improved cultivars and early cultivars (Fig. 7), which are descendants of orange carrots developed in Western Europe between the fifteenth and sixteenth centuries^50,51. These genetic signatures suggest a relatively recent contamination – not before the late Middle Ages – due to carrot residuals most likely associated to plant materials of Western European origin.

Another abundant crop among those identified in the TS samples is wheat and its close relatives. All cultivated polyploid forms originated and were domesticated in the Fertile Crescent and were subsequently introduced in the Mediterranean Basin from the Middle East. By the first century BC, with the emerging Roman Empire, wheat had become well-established in the region and recognized as an essential agricultural commodity. Additionally, the presence of exotic species introduced to Europe from Central and South-Eastern Asia, and North and South America, was noted. For example, maples, oranges and bananas were identified. Maples are thought to have originated in Eastern Asia, despite some evidence suggesting a North American origin⁴². Moreover, oranges, which are originally from the region encompassing Southern China and North-Eastern India, and bananas, whose earliest evidence of domestication is Papua New Guinea, were introduced to the Mediterranean Basin by the Muslim conquerors and Portuguese sailors, respectively, between the ninth and tenth centuries and fifteenth and sixteenth centuries^43,53. The identification of animal taxa posed a more intricate challenge owing to genetic similarities among closely related species (i.e. members of the Bovidae family). The presence of both wild and farm animals, including several livestock species and pets, indicates the enduring nature of agricultural and domestic practices prevalent in rural settings. The diversity of animals and plants at multiple taxonomic levels suggests that much of the environmental Shroud’s contamination occurred in recent centuries, particularly after the voyages of Marco Polo and Christopher Columbus. This hypothesis is consistent with and further corroborates our previous findings^3,14.

In this study, we also identified human mitochondrial DNA variants, which can be classified into one main haplogroup (K1a1b1a, which matches that of the collector), and those with diagnostic motifs of different lineages, such as H1b, an ancient late glacial marker common in Western Eurasia, and H33, which is particularly frequent among the Druze²² as previously documented by Barcaccia et al.¹⁴. Notably, the Druze population shares a common genetic ancestry with Jews and Cypriots, and has historically intermingled with other Levantine populations, including Palestinians and Syrians^21,22. Although additional single nucleotide polymorphisms (SNPs) were identified, it was not possible to reconstruct complete mitogenome haplotypes that could be confidently classified into any of the South Asian minor lineages (e.g., M39, M56, R7, and R8) identified in our previous study by analyzing mtDNA hypervariable segment I (HVS-I) from a different sample of the Shroud of Turin¹⁴. In that study, over 55.6% of the human mtDNA HVS-I reads were classified into Near Eastern lineages, while Western European lineages accounted for less than 5.6%. The presence of approximately 38.7% of Indian ethnic lineages could have resulted from historical interactions or the Romans importing linen from regions near the Indus Valley, associated with the term “Hindoyin” found in rabbinic texts¹⁵. Notably, the term “Shroud”, derived from the Greek "Sindôn" meaning fine linen, may be related to Sindh, a region renowned for its high-quality textiles. Historical evidence supports trade links between India and the Mediterranean, underscoring the significance of these textiles and inviting further exploration of ancient cultural interactions and trade practices^54,55. Indeed, the biblical scholar Levergne stated that the term "Sindôn" refers to a fabric of Indian origin, valued for its qualities and used for various and multiple purposes⁵⁶. In brief, a reappraisal of those outcomes from the analysis of the DNA traces found on the Shroud of Turin suggests the potentially extensive exposure of the cloth in the Mediterranean region and the possibility that the yarn was produced in India.

Overall, our prior and present findings provide valuable insights into the geographic origins of individuals who interacted with the Shroud throughout its historical journey across various regions, populations, and eras.

In this study, we sequenced and characterized genomic DNA traces found on the cloth edges, which were previously used for radiocarbon dating, and on other sections bearing the image of the Man of the Shroud. Molecular traces of contamination from individuals who handled the Shroud with limited protective precautions, including the 1978 sampling team in Turin, were expected. Our analysis revealed that the predominant mitochondrial lineage is characteristic of Ashkenazi Jews and matches the mtDNA haplotype of the late Prof. Baima Bollone, the collector, as also confirmed by kinship analyses on the entire genome. We also detected diagnostic variants of other haplogroups, such as the Western Eurasian haplogroups H2a2 and H1b, alongside H33, which is also found in Middle Eastern populations (i.e., Druze). The occurrence of an unusually high number of human mtDNA heteroplasmies, randomly distributed across the mitochondrial genome, and the coexistence of different mtDNA variants confirm that the Shroud came into contact with multiple individuals, thereby challenging the possibility of identifying the original DNA of the Shroud.

Intense human handling is confirmed by skin bacteria such as Cutibacterium and Staphylococcus. Halophilic archaea indicate salt preservation or saline storage conditions, but the presence of Debaryomyces hansenii and other fungi found in cheeses and hyper-saline waters raises further questions. Concerning the Shroud’s journey, the presence of red coral, livestock (e.g. chickens, cows, pigs, goats, sheep, rabbits, horses), and domestic cats and dogs suggest Mediterranean origins or transit through Mediterranean regions.

Plant DNA analyses revealed the predominance of cultivated species, which has historical implications. The concomitant and very abundant presence of both Daucus carota (domesticated carrot) and various Triticum species (cultivated wheat) suggests agricultural practices with both native and introduced crops. The carrot is the most prominent crop plant species identified in the Shroud. Wild populations of carrots with white roots have existed in Europe since more than two thousand years, as they were described by Pliny the Elder in the encyclopedic "Naturalis Historia" and used by Romans for medicinal purposes and food preparations (see Apicius, "De re culinaria" a collection of Roman cookery recipes). However, we demonstrated that the carrot DNA found on the Shroud is genetically more similar to early cultivars and improved cultivars. These cultivars were proven to descend from orange carrot varieties developed in Europe between the fifteenth and sixteenth centuries⁵². Among the other species identified, oranges and bananas, which are originally from Southern Asia and Eastern Oceania (China and Papua New Guinea), were introduced to the Mediterranean Basin by Muslim populations and Portuguese sailors in the late Middle Ages^43,53. Post-Columbus species, such as maize (domesticated in Mexico), Solanaceae crops (e.g. tomatoes, peppers, and potatoes), and peanuts (Fabaceae), which are crop plants native to Latin America, pose an additional significant conundrum about the Shroud’s antiquity, even if a later contamination cannot be ruled out. DNA traces from various species and regions, including the Middle East, the Mediterranean, Europe, America, and Asia, indicate that the Shroud was exposed to different environments and peoples. The prevalence of Mediterranean crops and the absence of typical Middle Eastern flora raise questions about the agricultural landscape when the Shroud was created or used as a burial cloth. Comparative analyses with other ancient textiles and artifacts could further illuminate cultural and historical interactions with plants and animals.

In conclusion, genetic and microbial evidence discloses a complex history of the Turin Shroud, reflecting interactions with a diverse array of individuals and exposure to various agricultural and environmental contexts. While some human mtDNA haplogroups correspond to those of individuals who handled and sampled the historical artifact in 1978, overall, the DNA results, derived from clean-room methodologies and metagenomic analyses supported by robust bioinformatics, suggest a mosaic of diverse genetic traces. The age of the Turin Shroud couldn’t be determined through metagenomics because this methodology is unable to provide any robust evidence supporting either a Medieval origin or a history dating back two millennia. Nevertheless, our findings constitute a novel and significant contribution to the field, thoroughly elucidating the biological traces left by centuries of social, cultural, and ecological engagement. Furthermore, radiocarbon dating of two distinct fabric threads collected from the reliquary confirms their use in mending and patching efforts to stabilize the fire-damaged linen cloth during two well-documented events in 1534 and 1694 CE. Collectively, our findings illuminate important aspects of the Shroud’s preservation history.

Addendum

In memory of Dr. Pierluigi Baima Bollone, coauthor of this publication, whose sudden departure has left our research groups at a loss for words.

As Pierluigi Baima Bollone passed away in Turin, Italy, on November 5, 2025, we wish to express our profound gratitude for his irreplaceable collaboration, invaluable contribution and unwavering dedication to the advancement of knowledge regarding the Turin Shroud.

Bibliographical notes: Pierluigi Baima Bollone was a prominent medical doctor specialized in hematopathology and a full professor of forensic medicine at the University of Turin. He founded and led the Department of Diagnosis and Prevention at Gradenigo Hospital. A prolific author of scientific publications, dissemination books and forensic medicine manuals, his works are widely adopted in academia. Renowned on both national and international stages, he produced influential writings in criminology and forensic medicine. He also served as Honorary President of the International Centre for Sindonology in Turin, the official institution devoted to the study of the Shroud, contributing extensively to its literature.

Sample collection

For the present study, twelve samples were retrieved from the 1978 Turin Shroud collection (Table 1 and Figure S1): seven glass tubes with linen cloth filaments, four jars with vacuum dust (one completely empty), and one microscope slide with a tiny biological fragment. This collection is entirely new compared to the one analyzed in Barcaccia et al.,¹⁴.

Permission and ethical approval

The permission to analyze the 1978 Official Sample Collections is provided in the SupplementaryDocument1.

All analyses were carried out in accordance with relevant guidelines and regulations. All research protocols were approved by the Ethical Committee Fondazione IRCCS Policlinico San Matteo (protocol number 0028298/22). Written informed consents were obtained from the collector and the operator, who also provided information about place of birth and genealogy.

Sequence data production

DNA was extracted from all available samples using a silica-based technique that allows for recovery of ancient/historical DNA even when molecules are very few and highly fragmented, as in Capodiferro et al.⁵⁷. After DNA extraction, DNA concentration and fragmentation were assessed by fluorometer and capillary electrophoresis, respectively. For samples with a DNA concentration higher than zero, shotgun genomic libraries were constructed following a custom double indexing protocol optimized for ancient DNA using partial uracil DNA glycosylase treatments (half-UDG), which partially retain the deaminated cytosines resulting from DNA damage⁵⁸.

Considering that 10 M reads are required for an accurate estimation of the endogenous content, the half-UDG genome libraries were multiplexed and sequenced on the Illumina platforms (NextSeq 550 and 2000) using a paired-end strategy. A present-day filament of linen was artificially contaminated by one of the handlers (operator) to validate the feasibility of our approach.

The modern DNA of the collector (Prof. Baima Bollone) was also extracted using the Maxwell® RSC Stabilized Saliva DNA Kit after collecting a biological sample with the IsoHelix GeneFiX™ Saliva DNA Collection Kit. For this modern sample, a genomic library was prepared with the Illumina DNA Prep Kit using the standard protocol and sequenced on the Illumina NextSeq 2000.

Sequence data analyses

The raw shotgun sequence data have been demultiplexed using bcl-convert v. 00.000.000.4.2.4⁵⁹. Raw fastq and adapter trimming have been performed using AdapterRemoval v. 2.3.2⁶⁰, discarding reads shorter than 30 bases after trimming and collapsing paired-end reads overlapping by at least 11 bases (parameters: –trimns –maxns 2 –trimqualities –minquality 2 –minlength 30 –collapse –minalignmentlength 11). Trimmed reads were mapped to the human reference genome hs37d5 using bwa v. 0.7.17-r1188⁶¹ with the aln/samse algorithms and parameters (-l 1024 -n 0.01 -o 2) optimized for ancient DNA⁶².

Although more complete and gapless human reference genomes are now widely used in modern human genomics, we used the hs37d5 reference, which is based on the GRCh37 assembly, because it remains the standard reference in the field of ancient human DNA. Indeed, the main curated database of ancient human genomic data⁶³ currently relies on hg19, which is also based on GRCh37, and even very recent ancient-DNA studies continue to use hs37d5/hg19 references⁶⁴. We selected hs37d5 specifically because the inclusion of decoy sequences can improve read alignment and reduce spurious mappings, which is particularly important for short and damaged ancient DNA fragments. After mapping, aligned reads were filtered for a minimum mapping quality of 20 and duplicates were removed using samtools v. 1.17⁶⁵ and DeDup v. 0.12.8⁶⁶, respectively. The two terminal bases of each read in the final BAM files were soft-clipped using trimBam from BamUtil v. 1.0.15⁶⁷, to mask potential aDNA damage patterns for downstream analyses. Statistics and aDNA damage patterns were evaluated with QualiMap v.2.2.2-dev (bamqc)⁶⁸, DamageProfiler v. 1.1⁶⁹, and PyDamage²⁹ (for reads mapped to assembled contigs, see “Taxonomic classification of the DNA sequences”). Genetic sex was evaluated following the two methods described in Skoglund et al.⁷⁰ and Mittnik et al.⁷¹. For the mtDNA, the trimmed reads were aligned to a modified version of the rCRS⁷², obtained using CircularGenerator v. 1.0 from CircularMapper⁷³, with bwa aln/samse and the same parameters as indicated above. RealignSAMFile v. 1.0 (from CircularMapper) was used to realign the reads to the original rCRS. BAM files were filtered in the same way as above. Mitochondrial DNA haplogroups were called using Haplocheck v. 1.3.2 with default options⁷⁴.

For a reliable haplogroup classification, ANGSD v 0.940-dirty⁷⁵ was also used to call a consensus mitochondrial sequence with the following parameters: -minQ 20 -setMinDepth 2 -minMapQ 30 -doFasta 2 -doCounts 1 -basesperLine 60. Consensus files were uploaded on HaploGrep v 3.2.1⁷⁶. The mitogenome data were also analyzed with Haplocheck v. 1.3.2⁷⁴, which was also used to construct schematic phylogenetic trees. All identified haplotypes therefore contain variants that distinguish them from the rCRS reference mtDNA sequence.

Contamination estimate

Contamination estimate on the mtDNA was obtained using contamMix v. 1.0–10⁷⁷. A consensus sequence from the aligned mtDNA reads was called with ANGSD v. 0.940-dirty⁷⁵ using the following parameters: –minMapQ 30 -minQ 20 -doCounts 1 -setMinDepth 3 -doFasta 2. The original mt-aligned reads were mapped to this consensus and used, together with a database containing 311 worldwide mitochondrial genomes, plus the consensus generated with ANGSD, to estimate potential mt contamination with contamMix (parameters: –nIter 50,000 –nChains 3 –trimBases 2). Nuclear contamination was estimated using hapConX⁷⁸ for those samples classified as male individuals.

PMDtools¹⁸ was used to assign PMD score to all reads mapped to the human genome and to extract endogenous ancient DNA sequences, separating them from contaminating modern sequences using postmortem degradation patterns.

Modern reference individual

The raw reads from the modern reference individual were processed using fastp v. 0.24.0⁷⁹ with parameters -g -x -c. Trimmed reads were mapped to the hs37d5 human reference using bwa-mem2⁸⁰. Duplicates were removed using picard-tools MarkDuplicates⁸¹ and final statistics were calculated using mosdepth v. 0.3.6 with default parameters⁸².

Kinship analysis

Kinship analyses were performed using the samples showing a sufficient average depth of coverage on the human genome, together with the modern reference individual. Three different analyses were performed using READ v. 2⁸³ and BREADR⁸⁴ and KIN⁸⁵. PileupCaller (from sequenceTools v. 1.5.3.2)⁸⁶ was used to call pseudohaploid genotype data on the 1240 K SNPs set to be used as input data for READ2 and BREADR.

Proteomics methods

Three TS samples (C/D12, B12, and R58) were extracted by adding 5% SDS / 50 mM TEAB buffer (20 µL to samples B12 and C/D12 and to two negative controls; 30 µL to R58). Extraction proceeded by alternating 30 min in a sonicator at 56 °C with 30 min on a thermomixer hotplate at 1,600 rpm, repeated over 5 h. Samples and negatives were then incubated overnight at 56 °C without shaking. The following day, five 1.4 mm ceramic beads, pre-cleaned with bleach, rinsed with methanol, and UV-irradiated overnight, were added to each tube. Mechanical homogenization was performed using a Precellys tissue lyser at 6500 rpm for 30 s. Lysates were centrifuged at 12,700 rpm (max speed) and supernatants transferred to protein low-bind tubes. For each sample, a final volume of 23 µL was prepared (adding extraction buffer as needed) for downstream S-Trap processing. Reduction was carried out by adding 1 µL of 60 mM DTT and incubating at 55 °C for 15 min. Alkylation followed by adding iodoacetamide (IAM) to a final concentration of 500 mM stock solution, incubating in the dark for 30 min at room temperature. Excess IAM was quenched by adding 1 µL DTT. To acidify the solution, 2.6 µL of 27.5% phosphoric acid was added, followed by 165 µL of S-Trap binding buffer consisting of 100 mM TEAB in 90% methanol. Acidified samples were loaded onto S-Trap micro spin columns and centrifuged at 4000 × g for 30 s. Each column was washed three times with 150 µL of binding buffer. Digestion was performed by adding 0.5 µg sequencing-grade trypsin in 50 mM TEAB directly onto the columns, followed by incubation for 2 h at 47 °C.

Peptides were eluted following the standard three-step S-Trap elution protocol: 40 µL of 50 mM TEAB, 40 µL of 0.2% formic acid in water and 40 µL of 50% acetonitrile / 0.2% formic acid. Each elution step included centrifugation at 4000 × g for 30 s, and eluates were pooled. Peptide solutions were dried overnight in a fume hood prior to MS analysis.

Peptide samples were analyzed by liquid chromatography–tandem mass spectrometry (LC–MS/MS) using a Thermo Scientific Orbitrap Astral mass spectrometer coupled to a Thermo Scientific Neo nano-UHPLC system via a nano-electrospray ionization source. Peptides were separated on a reversed-phase C18 analytical column (75 µm internal diameter, 25 cm length) operated at a nominal flow rate of 300 nL/min. Mobile phase A consisted of water and mobile phase B of acetonitrile, both containing 0.1% formic acid. Peptides were eluted using a linear gradient from 1 to 55% solvent B over 45 min, followed by a column wash at 99% B and re-equilibration, for a total run time of 52 min.

The mass spectrometer was operated in positive ion mode. Full MS survey scans were acquired in the Orbitrap analyzer over an m/z range of 350–1500 at a resolution of 240,000. Data-dependent MS/MS acquisition was performed with precursor ions selected for fragmentation using an isolation window of 1.5 m/z. Fragment ions were acquired in the Astral analyzer over an m/z range of 150–1800 using higher-energy collisional dissociation (HCD). Dynamic exclusion was enabled to minimize repeated sequencing of the same precursor ions. All other instrument parameters were set according to manufacturer recommendations and are reported in the supplementary methods.

The raw data were analyzed both Progenesis QI for Proteomics version 4.2. and FragPipe v.24.0 and MSFragger v.4.4.1.

For Progenesis QI for Proteomics, one experiment was conducted on the three samples only, and one separate experiment was conducted on the two negative controls, to identify proteins present in either the samples or in the negatives. After importing the data, all samples aligned against the selected reference with a similarity score greater than 90%. Peptide ions with charge 2–5 were kept, together with those with more than 2 isotopes. Overall, 3900 peptide ions were retained for the “samples” group, and 1359 for the “negatives” group. Peptides with a ranking 1–3 were exported in Mascot for identification. Search parameters were the following: semiTrypsin searches, with 2 missed cleavages allowed, fixed modification Carbamidomethyl (C) and variable modifications Deamidation (NQ) and Oxidation (M), peptide mass tolerance 10 ppm, fragment mass tolerance 0.02 Da, target FDR 1%. Results were further refined using Percolator. Peptide identified with a score smaller than 7 (for “samples”) and 12 (for “negatives”) were excluded from further analyses (as per Mascot peptide score distribution to indicate identity or extensive homology), as well as proteins identified by less than 2 unique peptides.

For FragPipe v.24.0 and MSFragger v.4.4.1. Specifically, all experiments were conducted independently, analyzing one sample per time. MSFragger settings were the following: precursor mass tolerance ± 10 ppm, fragment mass tolerance 0.02 Da, trypsin enzyme, 2 missed cleavages allowed, fixed modification Carbamidomethyl (C) and variable modifications Deamidation (NQ), Oxidation (M) and Acetylation at any protein N-terminal. The maximum number of variable modifications per peptide were set to 2, with 500 maximum combinations. MSBooster and PSM validation with Percolator were used to further refine searches. Proteins identified with no unique peptides were excluded from further analyses.

Taxonomic classification of the DNA sequences

The sequenced reads from the seven samples that underwent metagenomic analyses (Table 1) were processed to trim adapters and filter for quality using Trim Galore v0.6.7⁸⁷. Initial taxonomic classification of the reads was performed using Kraken v2.1.2⁸⁸ with default parameters for paired-end library reads. The database used was a combination of pre-built Kraken databases including archaea, bacteria, plasmids, human, fungi, viruses, protozoa, and plants genomes (PlusPFP database, version published 12 January 2024). An interactive visualization of the results, known as a Krona graph, was generated from the textual Kraken reports using KrakenTools (see Supplementary information, including Dataset S5)⁸⁹.

An additional classification of the reads, applicable only to microorganisms, was performed using the MetaPhlAn program²⁶. All reads assigned by Kraken to a taxon ID belonging to Archaea, Bacteria, or Fungi, or not assigned to any taxa, were extracted using TaxonKit v0.18.0⁹⁰, SeqKit v2.1.0⁹¹, and a custom awk command. The resulting paired-end FASTQ files were analyzed with MetaPhlAn v4.2.2 using the default settings, except for the parameter -t, which was set to “rel_ab_w_read_stats”.

Due to the absence of Metazoa records (except for Homo sapiens) in the Kraken database and the small number of Viridiplantae species included, Kraken2 cannot identify reads from these kingdoms. Therefore, a different approach was chosen. Raw reads were assembled into contigs for each sample using Megahit v1.2.9⁹² with the parameters –presets meta-large and –min-contig-len 100. Taxonomic classification of the resulting contigs was conducted using BLASTn v2.12.0⁹³ against the entire NCBI Nucleotide database⁹⁴, downloaded on December 6, 2024. The alignment parameters were: -evalue 1e-5, -max_target_seqs 5, -perc_identity 90, and -outfmt 6. The parameters qcov, qcovhsp, and staxids, were added to the standard fields for subsequent filtering and processing.

A preliminary filtering of the alignment results was performed to generate summary statistics. Alignments with query coverage below 90% were excluded, and for each contig, the best target based on E-value was selected. A custom Bash and Python3 script, utilizing TaxonKit v0.18.0⁹⁰, was used to calculate the number of contigs assigned to different taxa at each taxonomic rank.

To exclude the possibility that some nucleotide sequences, particularly those derived from GenBank, contained contaminant regions that could result in erroneous taxonomic assignments, the BLASTn results were screened against a database of contaminated sequences in both RefSeq and GenBank compiled by Steinegger and Salzberg in 2020⁹⁵. Sequence identifiers obtained from the BLASTn analysis were compared with those in the contamination database using a custom Python 3 script. This assessment revealed no evidence of contamination that could have affected the assignment of animal or plant taxa.

The alignments were then compared based on their highest-scoring HSP (high-scoring pair) for each target. This comparison aimed to determine whether the alignments were of similar significance. Alignments were deemed non-equivalent if any of the following criteria were not met: a maximum difference in percent query coverage of 3, a maximum difference percent identity of 3, a maximum difference in bit score of 10%, a maximum difference in E-value of three orders of magnitude. If one alignment outperformed all others, the contig was assigned to the corresponding taxonomic ID (taxid). If two or more alignments were equally probable, the contig was assigned to the lowest taxonomic node present in the genealogy of all target taxid. Contigs assigned to taxa above the family rank were discarded as too unspecific. This processing was implemented using a custom Bash and R script, incorporating TaxonKit⁹⁰ and the R package rentrez v1.2.3⁹⁶. All custom scripts are available upon request.

A final step was implemented to filter potential false positives. A manual interpretation of the tables containing all contigs assigned to a taxon was necessary to eliminate possible false positives. Two criteria were established: i) When multiple species within the same genus produced similar hits, or if the number of contigs assigned to the genus exceeded that of contigs assigned to any individual species within it, the contigs were assigned to an unidentified species within that genus (e.g., Musa spp.); ii) In cases where a substantial number of contigs identified a specific species, as well as the corresponding genus, subtribe, and tribe, all contigs were assigned to the species (e.g., Daucus carota). These reads likely represent broader, less specific regions that remain associated with the species, and it is prudent to accurately specify and categorize them accordingly.

This methodology was developed in response to findings related to monkeys, which likely yielded results associated with human DNA. Given our confidence regarding the presence of human DNA, it is highly improbable that various exotic monkey DNAs are present. We anticipate that in other instances, a small number of contigs will also identify species related to the primary target. This phenomenon may arise from the short contigs generated from degraded sequences, which can align, albeit infrequently, to incorrect targets when conserved regions are present.

Clustering of the samples

The analysis conducted on the seven samples was also applied to the modern reference individual (hereafter referred to as “control”), with minor variations in the steps involving Fastp, Megahit, and Kraken2. The same parameters were utilized, except for the use of single-end mode instead of paired-end mode, which was dictated by the sample’s nature.

To elucidate the differences among the samples, it was imperative to select a taxonomic rank that would minimize potential assignment errors between species or genera, thereby ensuring that reads assigned to that taxon could be regarded as accurate and reliable. The variation in classification practices across different kingdoms necessitated the choice of distinct taxonomic ranks corresponding to each kingdom: Archaea at the class level, Bacteria at the phylum level, and Metazoa and Fungi at the order level, while Viridiplantae was classified at the family level.

A matrix was constructed to display the number of contigs assigned to each defined taxonomic unit across the seven samples and the control. Subsequently, a principal component analysis (PCA) was performed using the R function “prcomp”, with the parameter scale = TRUE to prevent the Primates order from disproportionately influencing the variability. This clustering effectively represents the variance among the samples. However, it has the potential to overemphasize the significance of taxa represented by only a single contig in any given sample. To mitigate this issue, a second clustering analysis was conducted, wherein all taxa with a total number of contigs less than or equal to three were excluded. The number of contigs for each sample was then normalized by dividing each value by the maximum observed value, effectively rescaling all values to a range of 0 to 1 (max normalization by columns). This was followed by a row-wise max normalization to ensure that every taxon held equal weight in the analysis. Finally, the Euclidean distances between samples were calculated and subjected to hierarchical clustering using the R functions from the pheatmap package⁹⁷.

Genetic ancestry of carrot DNA contamination

To determine the genetic ancestry of the carrot DNA found on the Shroud, a comparative genomic analysis was performed against the Daucus carota reference genome⁵². Contigs from Shroud samples R58, AB1, B12, C10_gl, and C/D12 — identified as D. carota — were mapped independently to the reference genome. In parallel, Illumina reads (~ 10x) from 200 diverse carrot samples were aligned to the reference, and SNPs were identified using the GATK (Genome Analysis Toolkit) pipeline⁹⁸. The carrot samples represented five distinct populations as defined by Coe et al.⁵²: Early cultivars, Improved cultivars, Landrace A, Landrace B, and Wild. For each population, 40 accessions with an ancestry coefficient > 0.7 were selected. Both early and improved cultivars represent lineages descended from carrot cultivars developed in the fifteenth and sixteenth centuries. Landrace A and Landrace B consist of descendants from Central/Eastern Asia and Western/Southern Asia, respectively, exhibiting semi-domesticated phenotypes. Wild accessions represent undomesticated carrots from Europe. Because samples R58, AB1, B12, and C/D12 yielded < 3 uniquely mapped contigs, they were excluded from further analysis. In contrast, sample C10_gl provided 1787 uniquely mapped contigs and was used for all downstream analyses.

To estimate genetic similarity, Identical by State (IBS) counts were calculated using the overlapping SNPs between the carrot accessions and the C10_gl contigs. To ensure the reliability of the analysis and account for potential biases in contig distribution, a subsampling approach was employed: twenty contigs were randomly selected from the 1,787 available for C10_gl, with this process repeated for 100 iterations. The resulting non-IBS percentages were analyzed using t-tests and linear models in R⁹⁹ to determine if the genetic distances between C10_gl and the known phylogenetic groups were statistically significant. In this framework, lower non-IBS percentages denote higher genetic similarity, indicating a closer ancestral relationship. Plots were generated in R using ggplot2¹⁰⁰.

1. Jumper, E. J. et al. A comprehensive examination of the various stains and images on the Shroud of Turin. In Archaeological chemistry—III Vol. vol. 205 447–476 (American Chemical Society, 1984).

   Chapter  Google Scholar 

2. Adler, A. The orphaned manuscript: a gathering of publications on the Shroud of Turin (Effatà Editrice, 2014).

   Google Scholar 

3. Meacham, W. et al. The authentication of the Turin Shroud: An issue in archaeological epistemology [and comments and reply]. Curr. Anthropol. 24, 283–311 (1983).

   Article  Google Scholar 

4. Antonacci, M. The resurrection of the shroud: new scientific, medical, and archeological evidence (Evans M & Co, 2000).

   Google Scholar 

5. Fanti, G. & Basso, R. The Turin Shroud: optical research in the past, present and future Vol. 88 (Nova Science Publishers, 2008).

   Google Scholar 

6. Poulle, E. Les Sources de l’Histoire du Linceul de Turin. Revue Critique. Revue D’Histoire Ecclésiastique 104, 747–782 (2009).

   Article  Google Scholar 

7. Clari, R. de. La conquista di Costantinopoli (1198–1216) [di] Roberto di Clari. Studio critico, traduzione e note di Anna Maria Nada Patrone. (1972).

8. Rinaldi, P. Un documento probante sulla localizzazione in Atene della Santa Sindone dopo il saccheggio di Costantinopoli. in La Sindone scienza e fede, Atti del II Convegno nazionale di sindonologia 109–113 (1983).

9. Piana, A. The Shroud’s “Missing Years”. Shroud newsletter vol. 66 (2007).

10. Frale, B. ‘Redeat nobis quod sacrum est’ una lettera sulla presenza della Sindone in Atene all’indomani della quarta crociata. Aevum Anno 86, 589–641 (2012).

    Google Scholar 

11. Damon, P. E. et al. Radiocarbon dating of the Shroud of Turin. Nature 337, 611–615 (1989).

    Article  ADS  CAS  Google Scholar 

12. Casabianca, T., Marinelli, E., Pernagallo, G. & Torrisi, B. Radiocarbon dating of the Turin Shroud: New evidence from raw data. Archaeometry 61, 1223–1231 (2019).

    Article  ADS  CAS  Google Scholar 

13. Piana, A. Othon De La Roche and the Shroud. An Hypothesis between history and historiography. In Advances in the Turin Shroud Investigation (2014).

14. Barcaccia, G., Galla, G., Achilli, A., Olivieri, A. & Torroni, A. Uncovering the sources of DNA found on the Turin shroud. Sci. Rep. 5, 14484 (2015).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

15. Grossi, A. Jewish shrouds and funerary customs: a comparison with the Shroud of Turin. in 1–33 (Centro Español de Sindonologia (CES), Valencia, 2012).

16. di Riggi, G. Rapporto Sindone (1978/1987) - Cronache e commenti dei fatti accaduti dal 1978 al 1987 intorno alle ricerche scientifiche condotte sulla Sacra Sindone di Torino (3M Edizioni, 1988).

    Google Scholar 

17. The 1978 Scientific Examination. https://www.shroud.com/78exam.htm.

18. Skoglund, P. et al. Separating endogenous ancient DNA from modern day contamination in a Siberian Neandertal. Proc. Natl. Acad. Sci. U. S. A. 111, 2229–2234 (2014).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

19. Costa, M. D. et al. A substantial prehistoric European ancestry amongst Ashkenazi maternal lineages. Nat Commun 4, 2543 (2013).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

20. Cardinali, I. et al. Mitochondrial DNA footprints from Western Eurasia in modern Mongolia. Front. Genet. 12, 819337 (2021).

    Article  CAS  PubMed  Google Scholar 

21. Behar, D. M. et al. The genome-wide structure of the Jewish people. Nature 466, 238–242 (2010).

    Article  ADS  CAS  PubMed  Google Scholar 

22. Shlush, L. I. et al. The Druze: A population genetic refugium of the Near East. PLoS ONE 3, e2105 (2008).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

23. Tommasi, A. et al. Fifteen millennia of human mitogenome evolution in Sicily Sci. Adv. 11, eady1674 (2025).

    Google Scholar 

24. Nicolotti, A. The Shroud of Turin: the history and legends of the world’s most famous relic (R. A. Baylor University Press, 2019).

    Google Scholar 

25. Beldon Scott, J. Architecture for the Shroud: Relic and Ritual in Turin. The University of Chicago Press.

26. Blanco-Míguez, A. et al. Extending and improving metagenomic taxonomic profiling with uncharacterized species using MetaPhlAn 4. Nat. Biotechnol. 41, 1633–1644 (2023).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

27. Chin, C.-S et al. Genome of a thermophilic bacterium Geobacillus sp. TFV3 from Deception Island, Antarctica.

28. Byrd, A. L., Belkaid, Y. & Segre, J. A. The human skin microbiome. Nat. Rev. Microbiol. 16, 143–155 (2018).

    Article  CAS  PubMed  Google Scholar 

29. Borry, M., Hübner, A., Rohrlach, A. B. & Warinner, C. Pydamage: Automated ancient damage identification and estimation for contigs in ancient DNA de novo assembly. PeerJ 9, e11845 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

30. Govind, N. S. & Banaszak, A. T. Isolation and characterization of an autonomously replicating sequence (ARSD) from the marine yeast Debaryomyces hansenii. Mol Mar Biol Biotechnol 1, 215–218 (1992).

    CAS  PubMed  Google Scholar 

31. Estrada-de Los Santos, P., Vinuesa, P., Martínez-Aguilar, L., Hirsch, A. M. & Caballero-Mellado, J. Phylogenetic analysis of Burkholderia species by multilocus sequence analysis. Curr. Microbiol. 67, 51–60 (2013).

    Article  CAS  PubMed  Google Scholar 

32. Tahir, P., Ahmed, A. B., SaifulAzry, S. O. A. & Ahmed, Z. Retting process of some bast plant fibres and its effect on fibre quality: A review. BioResources 6, 5260–5281 (2011).

    Article  CAS  Google Scholar 

33. Melelli, A. et al. Lessons on textile history and fibre durability from a 4000-year-old Egyptian flax yarn. Nat. Plants 7, 1200–1206 (2021).

    Article  CAS  PubMed  Google Scholar 

34. Xu, Z. et al. Dandruff is associated with the conjoined interactions between host and microorganisms. Sci Rep 6, 24877 (2016).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

35. Sonnante, G., Pignone, D. & Hammer, K. The domestication of artichoke and cardoon: From Roman times to the genomic age. Ann. Bot. 100, 1095–1100 (2007).

    Article  PubMed  PubMed Central  Google Scholar 

36. Mahr, S. Cardoon, Cynara cardunculus. Wisconsin horticulture - division of extensons - university of Wisconsin (Madison, USA) https://hort.extension.wisc.edu/articles/cardoon-cynara-cardunculus/#:~:text=Cardoon%20is%20still%20cultivated%20for,was%20domesticated%20in%20ancient%20times (2025).

37. Tatke, P. & Waghmare, R. Antidiabetic plants with insulin mimetic activity. In Antidiabetic medicinal plants 491–513 (Elsevier, 2024). https://doi.org/10.1016/B978-0-323-95719-9.00009-4.

    Chapter  Google Scholar 

38. Grubben, G. & Denton, O. Plant resources of tropical africa vegetables 2 (PROTA Foundation/Backhuys Publishers/CTA, 2004).

    Google Scholar 

39. Delplancke, M. et al. Evolutionary history of almond tree domestication in the Mediterranean basin. Mol. Ecol. 22, 1092–1104 (2013).

    Article  CAS  PubMed  Google Scholar 

40. Kafkas, S. et al. Pistachio genomes provide insights into nut tree domestication and ZW sex chromosome evolution. Plant Commun. 4, 100497 (2023).

    Article  CAS  PubMed  Google Scholar 

41. Kislev, M. E., Hartmann, A. & Bar-Yosef, O. Early domesticated fig in the Jordan Valley. Science 312, 1372–1374 (2006).

    Article  ADS  CAS  PubMed  Google Scholar 

42. Li, J. et al. Maple phylogeny and biogeography inferred from phylogenomic data. J Sytematics Evol 57, 594–606 (2019).

    Article  ADS  Google Scholar 

43. Koeppel, D. Banana: the fate of the fruit that changed the world (Hudsen Street Press, 2008).

    Google Scholar 

44. Frei, M. Il passato della Sindone alla luce della palinologia. In La Sindone e la Scienza (ed. Coero-Borga, P.) 191–200 (Edizioni Paoline, 1979).

    Google Scholar 

45. Frei, M. Identificazione e classificazione dei nuovi pollini della Sindone. In La Sindone, Scienza e Fede 277–284 (CLUEB, 1983).

    Google Scholar 

46. Danin, A. & Baruch, J. Floristic indicators for the origin of the Shroud of Turin. In The Shroud of Turin/Unraveling the Mystery/Proceedings of the 1998 Dallas Symposium 202–214. Alexander Books, Alexander, North Carolina, USA, 2001.

47. Price, A. L. et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 38, 904–909 (2006).

    Article  CAS  PubMed  Google Scholar 

48. Cattaneo-Vietti, R. et al. An overexploited Italian treasure: Past and present distribution and exploitation of the precious red coral Corallium rubrum (L., 1758) (Cnidaria: Anthozoa). Ital. J. Zool. 83, 443–455 (2016).

    Article  CAS  Google Scholar 

49. Iorizzo, M. et al. Genetic structure and domestication of carrot (Daucus carota subsp. sativus) (Apiaceae). Am. J. Bot. 100, 930–938 (2013).

    Article  PubMed  Google Scholar 

50. Banga, O. Origin of the European cultivated carrot. Euphytica 6, 54–63 (1957).

    Article  Google Scholar 

51. Banga, O. Origin and distribution of the western cultivated carrot. Genetica Agraria 357–370 (1963).

52. Coe, K. et al. Population genomics identifies genetic signatures of carrot domestication and improvement and uncovers the origin of high-carotenoid orange carrots. Nat. Plants 9, 1643–1658 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

53. Langgut, D. The citrus route revealed: From Southeast Asia into the Mediterranean. HortScience 52, 814–822 (2017).

    Article  Google Scholar 

54. Young, G. K. K. Rome’s eastern trade: international commerce and imperial policy 31 BC–AD 305 (Routledge, 2011).

    Google Scholar 

55. ‘Cikai Korran came here and saw’: Visitors from India graffitied dozens of Egyptian tombs 2000 years ago | Live Science. https://www.livescience.com/archaeology/ancient-egyptians/cikai-korran-came-here-and-saw-visitors-from-india-graffitied-dozens-of-egyptian-tombs-2-000-years-ago.

56. Coero-Borga, P. La Sindone e La Scienza: Bilanci e Programmi—Atti Del II Congresso Internazionale Di Sindologia 1978 (Edizioni Paoline, 1978).

    Google Scholar 

57. Capodiferro, M. R. et al. Archaeogenomic distinctiveness of the Isthmo-Colombian area. Cell 184, 1706–1723 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

58. Rohland, N., Harney, E., Mallick, S., Nordenfelt, S. & Reich, D. Partial uracil–DNA–glycosylase treatment for screening of ancient DNA. Philos. Trans. R. Soc. B Biol. Sci. 370, 20130624 (2015).

    Article  Google Scholar 

59. Illumina. BLC Convert. Illumina (2023).

60. Schubert, M., Lindgreen, S. & Orlando, L. AdapterRemoval v2: Rapid adapter trimming, identification, and read merging. BMC Res. Notes 9, 88 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

61. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754–1760 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

62. Schubert, M. et al. Improving ancient DNA read mapping against modern reference genomes. BMC Genomics 13, 178 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

63. Mallick, S. et al. The Allen ancient DNA resource (AADR) a curated compendium of ancient human genomes. Sci. Data 11, 182 (2024).

    Article  PubMed  PubMed Central  Google Scholar 

64. Krettek, K.-L. et al. The shared genomic history of Middle- to Late-Holocene populations from the Southern Cone of South America. Curr. Biol. 36, 2382-2396.e14 (2026).

    Article  CAS  PubMed  Google Scholar 

65. Li, H. et al. The sequence alignment/map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).

    Article  PubMed  PubMed Central  Google Scholar 

66. Peltzer, A. et al. EAGER: Efficient ancient genome reconstruction. Genome Biol. 17, 60 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

67. Jun, G., Wing, M. K., Abecasis, G. R. & Kang, H. M. An efficient and scalable analysis framework for variant extraction and refinement from population-scale DNA sequence data. Genome Res. 25, 918–925 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

68. Okonechnikov, K., Conesa, A. & García-Alcalde, F. Qualimap 2: Advanced multi-sample quality control for high-throughput sequencing data. Bioinformatics 32, 292–294 (2016).

    Article  CAS  PubMed  Google Scholar 

69. Neukamm, J., Peltzer, A. & Nieselt, K. DamageProfiler: Fast damage pattern calculation for ancient DNA. Bioinformatics 37, 3652–3653 (2021).

    Article  CAS  PubMed  Google Scholar 

70. Skoglund, P., Storå, J., Götherström, A. & Jakobsson, M. Accurate sex identification of ancient human remains using DNA shotgun sequencing. J. Archaeol. Sci. 40, 4477–4482 (2013).

    Article  CAS  Google Scholar 

71. Mittnik, A., Wang, C.-C., Svoboda, J. & Krause, J. A molecular approach to the sexing of the triple burial at the upper Paleolithic site of Dolní Věstonice. PLoS ONE 11, e0163019 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

72. Andrews, R. M. et al. Reanalysis and revision of the Cambridge reference sequence for human mitochondrial DNA. Nat. Genet. 23, 147 (1999).

    Article  CAS  PubMed  Google Scholar 

73. Peltzer, A. Circular mapper. (2020).

74. Weissensteiner, H. et al. Contamination detection in sequencing studies using the mitochondrial phylogeny. Genom Res. 31, 309–316 (2021).

    Article  CAS  Google Scholar 

75. Korneliussen, T. S., Albrechtsen, A. & Nielsen, R. ANGSD: Analysis of next generation sequencing data. BMC Bioinform 15, 356 (2014).

    Article  Google Scholar 

76. Schönherr, S., Weissensteiner, H., Kronenberg, F. & Forer, L. Haplogrep 3: An interactive haplogroup classification and analysis platform. Nucleic Acids Res. 51, W263–W268 (2023).

    Article  PubMed  PubMed Central  Google Scholar 

77. Fu, Q. et al. A revised timescale for human evolution based on ancient mitochondrial genomes. Curr. Biol. 23, 553–559 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

78. Huang, Y. & Ringbauer, H. HapCon: Estimating contamination of ancient genomes by copying from reference haplotypes. Bioinformatics 38, 3768–3777 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

79. Chen, S., Zhou, Y., Chen, Y. & Gu, J. Fastp: An ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, i884–i890 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

80. Vasimuddin, Md., Misra, S., Li, H. & Aluru, S. Efficient architecture-aware acceleration of BWA-MEM for multicore systems. In 2019 IEEE International Parallel and Distributed Processing Symposium (IPDPS) 314–324. IEEE, Rio de Janeiro, Brazil, 2019. https://doi.org/10.1109/IPDPS.2019.00041.

81. Broad Institute. Picard toolkit. (2019).

82. Pedersen, B. S. & Quinlan, A. R. Mosdepth: Quick coverage calculation for genomes and exomes. Bioinformatics 34, 867–868 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

83. Alaçamlı, E. et al. READv2: Advanced and user-friendly detection of biological relatedness in archaeogenomics. Genome Biol. 25, 216 (2024).

    Article  PubMed  PubMed Central  Google Scholar 

84. Rohrlach, A. B., Tuke, J., Popli, D. & Haak, W. BREADR: an R package for the Bayesian estimation of genetic relatedness from low-coverage genotype data. Preprint at https://doi.org/10.1101/2023.04.17.537144 (2023)

85. Popli, D., Peyrégne, S. & Peter, B. M. KIN: A method to infer relatedness from low-coverage ancient DNA. Genome Biol 24, 10 (2023).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

86. Schiffels, S. SequenceTools. (2023).

87. Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet.J. 17, 10–12 (2011).

    Article  Google Scholar 

88. Wood, D. E., Lu, J. & Langmead, B. Improved metagenomic analysis with Kraken 2. Genome Biol 20, 257 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

89. Lu, J. et al. Metagenome analysis using the Kraken software suite. Nat. Protoc. 17, 2815–2839 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

90. Shen, W. & Ren, H. TaxonKit: A practical and efficient NCBI taxonomy toolkit. J. Genet. Genomics 48, 844–850 (2021).

    Article  PubMed  Google Scholar 

91. Shen, W., Sipos, B. & Zhao, L. SeqKit2: A Swiss army knife for sequence and alignment processing. iMeta 3, e191 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

92. Li, D., Liu, C.-M., Luo, R., Sadakane, K. & Lam, T.-W. MEGAHIT: An ultra-fast single-node solution for large and complex metagenomics assembly via succinct de Bruijn graph. Bioinformatics 31, 1674–1676 (2015).

    Article  CAS  PubMed  Google Scholar 

93. Camacho, C. et al. BLAST+: Architecture and applications. BMC Bioinform 10, 421 (2009).

    Article  Google Scholar 

94. Sayers, E. W. et al. Database resources of the national center for biotechnology information. Nucleic Acids Res. 50, D20–D26 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

95. Steinegger, M. & Salzberg, S. L. Terminating contamination: large-scale search identifies more than 2,000,000 contaminated entries in GenBank. Genome Biol 21, 115 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

96. Winter, D. J. Rentrez: An R package for the NCBI eUtils API. R J. 9, 520–526 (2017).

    Article  Google Scholar 

97. Kolde, R. pheatmap: Pretty heatmaps. (2018).

98. McKenna, A. et al. The genome analysis toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genom Res. 20, 1297–1303 (2010).

    Article  CAS  Google Scholar 

99. R Core Team. R: A language and environment for statistical computing. R foundation for statistical computing (2021).

100. Wickham, H. Data analysis. In ggplot2: Elegant graphics for data analysis (ed. Wickham, H.) 189–201 (Springer International Publishing, 2016). https://doi.org/10.1007/978-3-319-24277-4_9.

     Chapter  Google Scholar 

Download references

Author notes

1. Gianni Barcaccia, Nicola Rambaldi Migliore and Giovanni Gabelli have contributed equally to this work.

Authors and Affiliations

1. Laboratory of Genomics, Research Centre for Agriculture “Maurizio Borin”, Department DAFNAE, University of Padova, Campus of Agripolis, Legnaro, PD, Italy

   Gianni Barcaccia, Giovanni Gabelli, Fabio Palumbo & Andrea Squartini

2. Department of Biology and Biotechnology “L. Spallanzani”, University of Pavia, Pavia, Italy

   Nicola Rambaldi Migliore, Vincenzo Agostini, Elisabetta Moroni, Valeria Nicolini, Antonio Torroni & Alessandro Achilli

3. Plants for Human Health Institute, North Carolina State University, Kannapolis, NC, 28081, USA

   Liangliang Gao & Massimo Iorizzo

4. Laboratory of Criminalistic Science “Carlo Torre”, Department of Sciences of Public Health and Pediatrics, University of Torino, Turin, Italy

   Grazia Mattutino

5. Newcastle University Protein and Proteome Analysis (NUPPA) Facility, Medical School, Newcastle University, Newcastle Upon Tyne, UK

   Andrew Porter & Pawel Palmowski

6. Research Center for Field Archaeology and Forensic Taphonomy, University of Lancashire, Preston, UK

   Noemi Procopio

7. Southeastern Community College, West Burlington, IA, 52655, USA

   Ugo A. Perego

8. Department of Plant Sciences, College of Agriculture and Bioresources, University of Saskatchewan, Saskatoon, SK, S7N 5A8, Canada

   Timothy F. Sharbel

9. Professor Emeritus, University of Torino, Turin, Italy

   Pierluigi Baima Bollone

Contributions

Conceptualization: GB, AA; Data Curation: NRM, GG, AA; Formal Analysis: NRM, GG, VA, VN; Funding acquisition: AA, NRM, GB; Investigation: GB, NRM, GG, VA, FP, EM, AS, AA, MI, LG; Methodology: NRM, GG, VA, EM, AP, PP, NP; Project administration: GB, AA; Resources: GB, FP, GM, NP, PBB, AT, AA; Software: NRM, GG, VN; Visualization: GB, GG, VN; Supervision: GB, AT, AA; Validation: NRM, GG, AS; Writing – original draft: GB, AA; Writing – review & editing: GB, NRM, GG, FP, GM, UAP, TFS, PBB, AT, AS, AA, MI, LG.

Corresponding authors

Correspondence to Gianni Barcaccia or Alessandro Achilli.

Competing interests

The authors declare no competing interests.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions