Peptide Library v1.4: fragments from three additional theropod specimens, provenance tiers, and specimen chain-of-custody practice
1 Jurassic Meat Shenzhen Research Center, Shenzhen, China
* Corresponding author: Dr. Hao Lin · regulatory@jurassicmeat.com
Contents
Abstract
Library v1.4 adds the first peptide fragments from specimens other than the founding Kem Kem tooth (JM-SP-0001). Between 2014 and 2016, 27 candidate theropod specimens were considered under a new acquisition and intake procedure that assigns each candidate a provenance tier (A–C) before any destructive sampling. Ten candidates were declined on jurisdictional or documentary grounds, six were excluded after computed-tomography examination as composite or consolidated, and two were withdrawn because the holder would not permit sampling. Nine were sampled and three passed authentication: a Tier B tooth from the Kem Kem Group of Morocco, and a Tier A rib fragment and a Tier A tooth from the Hell Creek Formation of Montana, USA. Together they contribute 88 authenticated fragments, raising the library to 389 fragments and 3,050 unique residues; the remaining growth since v1.0 comes from three further extraction campaigns on JM-SP-0001 released in v1.1–v1.3. Coverage of the new specimens is low: of 2,402 alignment columns covered by any specimen, 80% are covered by one specimen only and 38 by all four. Fragments from the new specimens are shorter (median 9 residues) and less deamidated than those of JM-SP-0001, consistent with their younger age. All four specimens place on the stem lineage of Aves; relationships among them are not resolved. On a 19-taxon extant reference, the median per-fragment placement posterior reported by the v1.4 build is 0.47. JM-SP-0001 is retrospectively classified as Tier C and is retained.
Keywordspalaeoproteomicsspecimen provenancechain of custodyKem Kem GroupHell Creek FormationTheropodacontamination controlphylogenetic placement
JM-TR-2016-003 · Public summary of internal report · Shenzhen, November 2016
1. Summary
- First fragments from additional specimens. Three theropod specimens with documented provenance contribute 88 authenticated fragments. The library now holds 389 fragments and 3,050 unique residues from four specimens.
- Coverage is low. 80% of covered alignment positions are observed in one specimen only; 38 positions are observed in all four. Cross-specimen comparison is correspondingly limited.
- Provenance tiers are introduced. Every candidate specimen since 2015 has been assigned a tier (A–C) before sampling. New library specimens must be Tier A or B.
- The founding specimen is Tier C. JM-SP-0001 does not meet the standard introduced here. It is retained, and contributes 77% of library fragments.
- Placement: median per-fragment posterior 0.47 on a 19-taxon extant reference; all specimens place on the avian stem; relationships among them are unresolved.
2. Background
Library v1.0 (JM-TR-2014-002) consolidated the 214 authenticated fragments from a single isolated theropod tooth, JM-SP-0001, purchased in the Rissani Sunday market and described in JM-TR-2013-001. Maintenance releases v1.1 to v1.3 (Supplementary Table S1) added 87 further fragments from the same tooth, recovered in three extraction campaigns at the Shenzhen Research Center in 2014–2015, and expanded the extant reference set from 12 to 19 taxa. From v1.1, fragments have been held as one peptide file per specimen and aligned at build time by the automated bracketing module.
A single-specimen library has two obvious weaknesses. Every fragment inherits every uncertainty in that specimen's history, and there is no internal means of distinguishing a property of theropod proteins from a property of one tooth. More fragments from the same tooth address neither.
The obvious remedy is more specimens. It is not straightforward. Fossil material containing endogenous protein is rare, and most theropod material available outside institutional collections is commercially traded. Commercial material is frequently composite, restored or consolidated, often with adhesives of animal origin; its locality data are often absent or unverifiable; and the legality of export varies by country of origin. Protein recovered from any fossil must be presumed contaminant until shown otherwise (Buckley et al. 2008), and the mineral matrix may itself be the condition for genuine survival (Demarchi et al. 2016), which makes the history of that matrix a scientific variable rather than a curatorial detail.
Between 2014 and 2016 we therefore established a specimen acquisition and intake procedure (SOP-SP-003) before acquiring further material, and applied it to every candidate considered. This report describes that procedure, the three specimens that passed it, the fragments they contributed, and Library v1.4.
3. Materials and methods
3.1 Provenance tiers
Each candidate is assigned a provenance tier at intake (Table 1), on the basis of documentation supplied and verified. Tier is assigned before any examination that might bias it. A tier can be lowered by subsequent findings; it cannot be raised without new documentation. Each candidate receives a specimen register number at intake (JM-SP-0002 to JM-SP-0028 in this report; JM-SP-0001 is the founding specimen). Numbers are not reused, and a registered specimen enters the library only on acceptance.
| Tier | Requirements | Eligible for library? |
|---|---|---|
| A | Locality to named site or coordinates, with formation; collection date; documented legal title from collection (landowner agreement or institutional accession); export and import documentation; unbroken custody record from collection or from first documented holder | Yes |
| B | Formation and region; documented legal title; export permit from the country of origin and import documentation; custody record from export onward | Yes |
| C | Attestation of the purchaser or holder only | No (new specimens) |
| — | Material from a jurisdiction in which commercial export of vertebrate fossils is prohibited | Declined without examination |
3.2 Intake and chain of custody
On receipt, each candidate is logged, photographed on all faces, weighed, sealed in a tamper-evident bag with a numbered seal, and stored at 4 °C in a dedicated specimen cabinet with restricted access. Every subsequent event (unsealing, examination, sampling, resealing, transfer) is recorded in the paper custody log, with operator, date, time, seal numbers and purpose, and transcribed weekly to the specimen register. The three accepted specimens accumulated between 12 and 19 custody events between intake and the v1.4 build.
All candidates passing documentary review were examined by micro-computed tomography before sampling, to identify composite construction (bonded roots, inserted crown fragments, filled cracks) and visible consolidant or adhesive. Composite specimens were excluded. Specimens showing consolidant were excluded unless sampling more than 5 mm from any consolidated region was possible.
3.3 Laboratory environment
Sampling and extraction were performed in the Shenzhen Research Center's dedicated ancient-protein clean room, commissioned in early 2015: positive pressure with HEPA-filtered supply, UV irradiation between sessions, full-body coveralls, double gloving, and single-use plasticware. No modern vertebrate tissue is handled in the clean room. Modern positive controls are prepared in a separate laboratory on a different floor.
3.4 Sampling, extraction and LC-MS/MS
Sampling and extraction followed JM-TR-2013-001 (§3.4) with three changes: surface removal was increased to 1.0 mm; two independent subsamples were taken from each specimen, on different days, by different analysts; and each extraction batch included two extraction blanks. Acid-soluble and acid-insoluble fractions, digested and undigested halves, and (for teeth) separate enamel extracts were processed as before. For the rib fragment, cortical bone was sampled from a freshly cut internal face after removal of the periosteal surface. Instrument parameters are given in Table 2.
| Parameter | Setting |
|---|---|
| Column | 75 µm × 50 cm, C18, 2 µm particles; heated to 50 °C |
| Gradient | 3–32% acetonitrile in 0.1% formic acid over 120 min; 250 nL/min |
| MS1 | Orbitrap-class; resolution 120,000 at m/z 200; m/z 350–1,500; AGC 4×10⁵ |
| MS2 | Top-15 data-dependent HCD; resolution 30,000; AGC 5×10⁴; max. injection 100 ms |
| Isolation window | 1.4 m/z |
| Normalised collision energy | 28 |
| Charge states | 2+ to 6+ |
| Dynamic exclusion | 30 s |
| Injection order | Randomised across specimens within batch; solvent blank before and after every specimen injection |
| Solvent (injection) blanks | 64 |
| Extraction blanks | 18 |
JM-SP-0001 extracts from Campaigns D–F (v1.1–v1.3) were last analysed on this instrument in October 2015. Before the first candidate injection in 2016 the analytical column was replaced and 24 consecutive solvent blanks were acquired; none contained a JM-SP-0001 library sequence.
3.5 Database search and authentication
Searches used MaxQuant (Cox and Mann 2008) with the v1.3 reference database (19 taxa), the contamination database extended to 312 sequences, and reversed decoys; parameters and de novo procedure were as in JM-TR-2013-001 (§3.6). The extension added collagen and gelatin sequences from the animal sources commonly used in hide and bone glues, after two candidates' extracts were dominated by them (§4.1).
Authentication criteria were as in JM-TR-2013-001 (§3.8), with criterion 3 (observation in at least two independent campaigns) replaced by observation in both independent subsamples. One criterion was added:
- Cross-specimen check. Any peptide observed in two or more specimens processed in the same batch was required to be present in at least one extraction from a batch not containing the other specimen.
3.6 Reference set, bracketing and placement
The extant reference set at v1.4 comprises 19 taxa (Supplementary Table S2). Fragments are aligned and applied as bracketing constraints at build time by the automated bracketing module introduced in v1.1. Placement follows JM-TR-2014-002 (§3.7) and is computed by the placement pipeline, which reads fragments directly from the library and does not depend on the bracketing step. For the four-specimen analysis (§4.5), specimens were placed jointly on the fixed extant reference topology, with relationships among fossil specimens estimated and all extant relationships held constant; ML support was estimated from 1,000 bootstrap replicates.
4. Results
4.1 Candidate screening
Twenty-seven candidates were considered (Table 3, Fig. 1). Four were offered from jurisdictions in which commercial export of vertebrate fossils is prohibited and were declined without examination. Six were declined because documentation was insufficient for Tier B; in five of these the documentation consisted of a receipt. Six were excluded after CT examination: four showed bonded roots from a different tooth, one a crown assembled from three fragments, and one was substantially infilled with pigmented resin. Two holders would not permit destructive sampling.
| Outcome | Kem Kem Group (Morocco) | Hell Creek Formation (USA) | Other | Total |
|---|---|---|---|---|
| Candidates considered | 14 | 9 | 4 | 27 |
| Declined: export prohibited in country of origin | — | — | 4 | 4 |
| Declined: documentation below Tier B | 5 | 1 | — | 6 |
| Excluded at CT: composite or consolidated | 5 | 1 | — | 6 |
| Withdrawn: sampling not permitted | 1 | 1 | — | 2 |
| Sampled | 3 | 6 | — | 9 |
| Failed authentication | 2 | 2 | — | 4 |
| Insufficient yield (<15 authenticated fragments) | — | 2 | — | 2 |
| Accepted into library | 1 | 2 | — | 3 |
Of the four authentication failures, two (JM-SP-0006 and JM-SP-0016, both Kem Kem teeth) yielded extracts in which more than 90% of PSMs were assigned to bovine or porcine collagen with low deamidation, consistent with penetration of an animal-glue adhesive through microcracks below the CT resolution used. One (JM-SP-0008, Hell Creek) failed replication between subsamples. One (JM-SP-0023, Hell Creek) produced candidate sequences also present in extraction blanks of the same batch, and was not re-extracted.
4.2 Accepted specimens
The three accepted specimens are described in Table 4.
- JM-SP-0011 is a Kem Kem tooth crown attributed on morphology to Abelisauridae, purchased in 2015 from a licensed exporter in Erfoud with a Moroccan export permit. Its locality is recorded as "Kem Kem, near Taouz." It is Tier B.
- JM-SP-0014 is a rib fragment from the Hell Creek Formation (Maastrichtian) of Montana, attributed to Theropoda (cf. Tyrannosauridae) by the lending collection. It is held on institutional loan under a written agreement permitting destructive sampling of up to 250 mg, with collection locality, date and accession history documented. It is Tier A.
- JM-SP-0019 is a tooth from the Hell Creek Formation attributed to Dromaeosauridae, collected in 2013 by a commercial collector on private ranch land under a written landowner agreement, with site coordinates recorded, and purchased in 2015 from the collector's dealer. It is Tier A. Before purchase it was held in the dealer's inventory for approximately two years, in a shared tray with other teeth, and was handled by prospective buyers at two trade shows.
| Specimen | Element | Formation (age) | Attribution | Tier | Fragments | Unique residues | Median length | Q / N deamidation | Median posterior |
|---|---|---|---|---|---|---|---|---|---|
| JM-SP-0001 | Tooth | Kem Kem (Cenomanian) | Carcharodontosauridae indet. | C | 301 | 2,190 | 11 | 0.74 / 0.89 | 0.48 |
| JM-SP-0011 | Tooth | Kem Kem (Cenomanian) | Abelisauridae indet. | B | 31 | 312 | 9 | 0.71 / 0.86 | 0.42 |
| JM-SP-0014 | Rib | Hell Creek (Maastrichtian) | Theropoda (cf. Tyrannosauridae) | A | 38 | 371 | 10 | 0.64 / 0.82 | 0.51 |
| JM-SP-0019 | Tooth | Hell Creek (Maastrichtian) | Dromaeosauridae indet. | A | 19 | 177 | 8 | 0.62 / 0.80 | 0.38 |
| Library v1.4 | 389 | 3,050 | 10 | 0.47 |
| Specimen | Material sampled (mg) | MS2 spectra | PSMs (1% FDR) | Unique sequences (1% FDR) | Authenticated fragments |
|---|---|---|---|---|---|
| JM-SP-0011 | 142 (crown dentine 124, enamel 18) | 86,920 | 1,184 | 133 | 31 |
| JM-SP-0014 | 210 (cortical bone) | 94,310 | 1,436 | 157 | 38 |
| JM-SP-0019 | 61 (crown dentine 52, enamel 9) | 45,880 | 602 | 71 | 19 |
| Extraction blanks (n = 18) | — | 401,770 | 512 | 64 | 0 |
JM-SP-0019 is small. The 61 mg sampled represents approximately 38% of its crown mass. The remainder is retained.
4.3 Authentication
Deamidation in the accepted specimens is high and clearly separated from that of the failed candidates, the blanks, and modern controls (Fig. 2). The two insufficient-yield candidates show deamidation comparable to the accepted specimens; they failed on quantity, not character. The Hell Creek specimens are less deamidated than the Kem Kem specimens, consistent with their being some 25–30 million years younger, although deamidation depends heavily on burial environment and is not used here as a measure of age.
Source data
Series Label Deamidated fraction, Gln (Q) Deamidated fraction, Asn (N) Accepted JM-SP-0011 0.71 0.86 Accepted JM-SP-0014 0.64 0.82 Accepted JM-SP-0019 0.62 0.8 Insufficient yield JM-SP-0021 0.58 0.77 Insufficient yield JM-SP-0025 0.64 0.79 Failed authentication JM-SP-0006 0.14 0.22 Failed authentication JM-SP-0008 0.21 0.31 Failed authentication JM-SP-0016 0.09 0.17 Failed authentication JM-SP-0023 0.18 0.27 Reference and controls JM-SP-0001 0.74 0.89 Reference and controls Human keratin, blanks 0.06 0.11 Reference and controls Modern bovine bone 0.04 0.09
No candidate-endogenous PSMs were observed in any of the 18 extraction blanks. The cross-specimen check (criterion 6) removed 7 candidate peptides, all shared between JM-SP-0014 and JM-SP-0019, which had been extracted in the same batch; on re-extraction in separate batches, 5 of the 7 were found in only one of the two specimens. Nine of the 31 JM-SP-0011 fragments are identical in sequence to JM-SP-0001 fragments. All nine lie in conserved regions of the COL1A1 helix. Identity at conserved collagen positions between two Cenomanian theropods from the same formation is unremarkable.
4.4 Composition, length and coverage
Collagen type I accounts for 70% of library fragments (Fig. 3). JM-SP-0014, a rib, contributes no enamel fragments; its non-collagenous fragments derive from osteocalcin and secreted phosphoprotein 1. The enamel fraction of JM-SP-0001 increased from 29 to 74 fragments over v1.1–v1.3, following an enamel-targeted sampling campaign (Campaign E) in the clean room in 2015.
Fragments from the new specimens are shorter than those from JM-SP-0001 (median 9 versus 11 residues; Fig. 4), attributable in part to lower yields and in part to the higher proportion recovered from undigested extracts (67%, against 58% for JM-SP-0001). Fragments shorter than 8 residues make up 19% of the new set, against 14% for JM-SP-0001.
Source data
Fragment length (residues) Fragments 6–8 17 8–10 28 10–12 17 12–14 12 14–16 7 16–20 5 20–28 2
The new specimens add 860 specimen-residues to the library but only 212 alignment columns not already covered by JM-SP-0001. Coverage across specimens is therefore thin (Table 6, Fig. 5). Of the 2,402 alignment columns covered by at least one specimen, 1,927 (80%) are covered by exactly one. Thirty-eight columns, all in the triple-helical domain of COL1A1, are covered by all four specimens.
| Coverage depth (specimens) | Alignment columns | Contribution to unique residues |
|---|---|---|
| Exactly 1 | 1,927 | 1,927 |
| Exactly 2 | 340 | 680 |
| Exactly 3 | 97 | 291 |
| All 4 | 38 | 152 |
| Covered by ≥1 | 2,402 | 3,050 |
Source data
Specimen 1 2 3 4 5 6 7 8 9 10 JM-SP-0001 0.44 0.92 0.95 0.91 0.83 0.42 0.74 0.78 0.69 0.33 JM-SP-0011 0.02 0.18 0.29 0.24 0.12 0.01 0.09 0.11 0.05 0 JM-SP-0014 0.05 0.27 0.36 0.31 0.19 0.04 0.17 0.21 0.12 0.02 JM-SP-0019 0 0.08 0.14 0.11 0.04 0 0.03 0.05 0.01 0
4.5 Placement
The v1.4 build reports a median per-fragment placement posterior of 0.47, computed on the 276 diagnostic fragments without a placement override, against 0.41 at v1.0. Most of the improvement is attributable to the reference expansion in v1.1–v1.3 rather than to the new specimens. The dependence of placement confidence on fragment length persists (Fig. 6); fragments of 6–7 residues remain high-variance, with interquartile ranges spanning most of the probability scale.
Source data
Fragment length (residues) v1.0 v1.4 v1.4 interval 6 0.19 0.21 0.1–0.33 7 0.24 0.28 0.15–0.41 8 0.29 0.37 0.22–0.5 9 0.34 0.45 0.3–0.58 10 0.38 0.52 0.38–0.64 12 0.44 0.61 0.47–0.72 14 0.5 0.68 0.55–0.78 16 0.55 0.72 0.6–0.82 20 0.62 0.79 0.68–0.87 24 0.67 0.83 0.73–0.9
Placed alone, each specimen places on the stem lineage of Aves: JM-SP-0001, posterior 0.94; JM-SP-0011, 0.87; JM-SP-0014, 0.90; JM-SP-0019, 0.68, with 0.24 for placement within crown Aves. The weaker result for JM-SP-0019 reflects its 19 fragments and short median length.
In the joint analysis (Fig. 7), JM-SP-0019 is recovered as the fossil specimen closest to crown Aves (bootstrap 51%), followed by JM-SP-0014 (39%), with the two Kem Kem specimens grouped together (36%). The first two relationships are what morphology would predict for a dromaeosaurid and a tyrannosaurid. The third is not: abelisaurids and carcharodontosaurids are not expected to be each other's closest relatives. All three nodes are weakly supported, rest substantially on the 38 columns shared by all specimens, and are not interpreted.
Source data
Taxon Highlighted Placement posterior JM-SP-0019 yes 0.51 Struthio Tinamus Gallus Meleagris Anas Columba Calypte Haliaeetus Falco Melopsittacus Corvus Taeniopygia JM-SP-0014 yes 0.39 JM-SP-0001 yes 0.36 JM-SP-0011 yes 0.36 Alligator Caiman Crocodylus Gavialis Chrysemys Pelodiscus Anolis ((((((JM-SP-0019,((Struthio,Tinamus),(((Gallus,Meleagris),Anas),(Columba,(Calypte,(Haliaeetus,(Falco,(Melopsittacus,(Corvus,Taeniopygia))))))))),JM-SP-0014),(JM-SP-0001,JM-SP-0011)),((Alligator,Caiman),(Crocodylus,Gavialis))),(Chrysemys,Pelodiscus)),Anolis);
5. Discussion
Library v1.4 is the first version in which the library is not a description of one tooth. That is its main contribution, and its main limitation is that it is only barely not a description of one tooth: JM-SP-0001 still supplies 77% of fragments and 72% of unique residues, covers 91% of all covered alignment columns, and contributed more fragments in v1.1–v1.3 alone than the three new specimens together.
The yield of the acquisition procedure was low: three accepted specimens from 27 candidates, over approximately two years. We consider the procedure's exclusions to be its most useful output. Two candidates that passed documentary review and CT examination would have entered the library on their database search results alone, had deamidation and the extended contamination database not identified their extracts as dominated by modern adhesive collagen. Two further specimens extracted in the same batch shared seven peptides, five of which were found in only one of the two when the specimens were extracted apart. Neither finding would have been possible with the controls used for JM-SP-0001, which was extracted alone and had no comparator.
That observation applies with some force to JM-SP-0001 itself. Under the tiers introduced here it is Tier C: its locality, collection date, and history before 2011 are unrecorded, and its documentation consists of the recollection of its purchaser. An attempt in 2016 to locate the dealer from whom it was purchased was not successful. Had it been offered as a candidate in 2015, it would have been declined at documentary review and not examined. We have retained it. Its fragments were authenticated under the most rigorous controls available at the time, have been reproduced across six campaigns in three laboratories, and behave in every respect examined here like fragments from specimens that do meet the standard. The decision was taken by the curators with the Chief Scientific Officer and recorded as a foundational-specimen provision. It applies to JM-SP-0001 only.
Placement continues to be limited more by fragment length than by reference density or specimen count (Fig. 6), and the joint placement of the four specimens should be read in that light. That JM-SP-0019, a dromaeosaurid tooth, is the specimen closest to birds is consistent with expectation and is supported by 19 fragments, mostly short. We note it, and draw nothing from it.
6. Limitations and known issues
- Low coverage. 80% of covered alignment columns are observed in one specimen only. Cross-specimen inference is possible at 38 columns.
- Foundational specimen is Tier C. JM-SP-0001, contributing 301 of 389 fragments, would not be accepted under current practice. It is retained under a foundational-specimen provision.
- Dealer-inventory exposure. JM-SP-0019 was held for approximately two years in a shared dealer tray and handled at two trade shows. Surface removal was increased to 1.0 mm in response.
- Locality precision. JM-SP-0011's locality is "near Taouz." Taouz is a village.
- Morphological attributions are to family level or above, from an isolated tooth or a rib fragment. The attribution of JM-SP-0014 is that of the lending collection and was not re-examined.
- Adhesive collagen. CT at the resolution used did not detect animal-glue penetration in two candidates. Accepted specimens showed no excess of bovine or porcine collagen, but the detection limit for low-level adhesive contribution has not been established.
- Short fragments. 19% of new fragments are shorter than 8 residues. Their placement is high-variance.
- Bracketing constraint count. The v1.4 bracketing log reports "constraints applied: 248" for a library of 389 fragments. This figure has increased with every build since v1.1 and is not reconciled against the library fragment count as part of the release procedure.
- Provenance metadata location. Provenance tier and custody records are held on paper and in the specimen register, not in the library. Library entries carry the specimen ID only.
- Override flags. The 37 manual override flags applied at v1.0 remain in force. Four further flags were added in v1.2 (Table S1). None has been reviewed since it was applied.
7. Changes in Library v1.4
As published in the Peptide Library Release Notes:
- Added first fragments from additional specimens, acquired with documented provenance. Low coverage.
In detail:
- New specimens
- JM-SP-0011 (Kem Kem Group, tooth, Tier B); JM-SP-0014 (Hell Creek Formation, rib, Tier A); JM-SP-0019 (Hell Creek Formation, tooth, Tier A).
- Fragments
- 389 (301 at v1.3, plus 31, 38 and 19 from the new specimens). Unique residues 3,050 (v1.3: 2,190).
- References
- 19 extant taxa (unchanged from v1.3).
- Authentication
- subsample-level replication for new specimens; cross-specimen batch check added; contamination database extended to 312 sequences.
- Placement
- median per-fragment posterior 0.47 (v1.0: 0.41).
- Specimen policy
- new fossil-derived entries require a Tier A or B specimen. JM-SP-0001 (Tier C) retained under a foundational-specimen provision.
8. Data availability
Library contents are confidential. Library v1.4 and all prior versions are archived and remain available to authorized research staff. The specimen register, custody logs, CT data and acquisition documentation are retained by Jurassic Meat. JM-SP-0001, JM-SP-0011 and JM-SP-0019, and retained material from all sampled candidates, are stored at the Shenzhen Research Center; JM-SP-0014 is held under the terms of its loan agreement. No library content is available to third parties.
9. Acknowledgements
We thank the lending collection, licensed exporters, collectors and landowners who supplied specimens and provenance documentation; M. Oyelaran for the commissioning and operation of the ancient-protein clean room; R. Venkataraman for review of the joint placement analysis; and a dealer in the Rissani Sunday market for assistance with specimen provenance.
Funding. This work was funded internally.
Competing interests. All authors are employees of Jurassic Meat and may hold equity.
Ethics. Reviewed by the Ethics & Bioresponsibility Council; no concerns raised. Acquisition of each specimen was reviewed for compliance with the export and import requirements of the jurisdictions concerned.
10. References
- Bertazzo S, et al. (2015). Fibres and cellular structures preserved in 75-million-year-old dinosaur specimens. Nature Communications 6:7352.
- Buckley M, et al. (2008). Comment on "Protein sequences from mastodon and Tyrannosaurus rex revealed by mass spectrometry." Science 319:33.
- Cox J, Mann M (2008). MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nature Biotechnology 26:1367–1372.
- Demarchi B, et al. (2016). Protein sequences bound to mineral surfaces persist into deep time. eLife 5:e17092.
- Schweitzer MH, et al. (2009). Biomolecular characterization and protein sequences of the Campanian hadrosaur B. canadensis. Science 324:626–631.
- Witmer LM (1995). The extant phylogenetic bracket and the importance of reconstructing soft tissues in fossils. In: Thomason JJ (ed.), Functional Morphology in Vertebrate Paleontology. Cambridge University Press, 19–33.
- JM-TR-2013-001. Endogenous peptide fragments recovered from an isolated Kem Kem Group theropod tooth.
- JM-TR-2014-002. Peptide Library v1.0.
Supplementary tables
| Version | Date | Fragments | Unique residues | Reference taxa | Principal changes |
|---|---|---|---|---|---|
| 1.0 | 2014-03 | 214 | 1,880 | 12 | First structured release (JM-TR-2014-002); manual bracketing |
| 1.1 | 2014-09 | 238 | 1,990 | 12 | JM-SP-0001 Campaign D (Shenzhen, 2014): +24 fragments; per-specimen peptide files; automated bracketing module introduced; crocodilian references refreshed |
| 1.2 | 2015-04 | 263 | 2,070 | 15 | JM-SP-0001 Campaign E (enamel-targeted, clean room): +25 fragments; added Melopsittacus, Haliaeetus, Calypte; SOP-PL-001 rev. 2; 4 override flags added |
| 1.3 | 2015-12 | 301 | 2,190 | 19 | JM-SP-0001 Campaign F: +38 fragments; added Corvus, Tinamus, Caiman, Pelodiscus |
| 1.4 | 2016-11 | 389 | 3,050 | 19 | JM-SP-0011, JM-SP-0014, JM-SP-0019: +88 fragments (this report) |
| Group | Taxa |
|---|---|
| Palaeognaths (2) | Struthio camelus, Tinamus guttatus |
| Galloanserae (3) | Gallus gallus, Meleagris gallopavo, Anas platyrhynchos |
| Neoaves (7) | Columba livia, Calypte anna, Haliaeetus leucocephalus, Falco peregrinus, Melopsittacus undulatus, Corvus brachyrhynchos, Taeniopygia guttata |
| Crocodylia (4) | Alligator mississippiensis, Caiman crocodilus, Crocodylus porosus, Gavialis gangeticus |
| Testudines (2) | Chrysemys picta, Pelodiscus sinensis |
| Squamata (1) | Anolis carolinensis |