Tree-free molecular clock calibration across millions of pathogen genomes — Manuscript, Streaming Scripts, Sequence ID Manifests, and Complete Replication Suite
Authors/Creators
- 1. Institute for Genomics and Evolutionary Medicine, Temple University, Philadelphia, PA, USA; Department of Biology, Temple University, Philadelphia, PA, USA
- 2. Institute of Infectious Disease and Molecular Medicine, University of Cape Town, Cape Town, South Africa
- 3. Department of Biochemistry and Molecular Genetics, University of Louisville, Louisville, KY, USA
- 4. Baker Institute for Animal Health, Cornell University, Ithaca, NY, USA
- 5. Department of Medicine, University of California San Diego, La Jolla, CA, USA
- 6. Institute for Genomics and Evolutionary Medicine, Temple University, Philadelphia, PA, USA
- 7. Department of Biochemistry and Molecular Biology, The Pennsylvania State University, University Park, PA, USA
Description
Complete replication package, datasets, streaming ingestion scripts, sequence ID manifests, deterministic verification suite, and autonomous agent replication playbooks for the manuscript:
Tree-free molecular clock calibration across millions of pathogen genomes
Modern pathogen sequencing generates millions of genomes, yet molecular clock calibration still requires reconstructing a phylogenetic tree—forcing pipelines to subsample over 99% of data or expend weeks of compute. Here we show that tree reconstruction is unnecessary to calibrate a molecular clock: dating a population requires only anchoring the ancestral root and weighting shared ancestry across sublineages. We introduce ChronAeon, which executes both operations in continuous distance space via a time-weighted profile root, foundation-model attention (HyphAeon), and spectral partitioning (AutoClock). Across empirical and synthetic benchmarks, ChronAeon matches Bayesian accuracy in seconds while flagging uninformative regimes, and streams millions of genomes to detect emerging variant clocks weeks before global dominance.
Contents of this deposit (v2.0.0):
chronaeon_nature_genetics.pdf&chronaeon_supporting_info.pdf: Main manuscript and Supplementary Notes 1–6, Figures S1–S4, and Tables S1–S3.chronaeon_nature_genetics_replication_suite.tar.gz(162 MB): Complete self-contained replication suite including thechronaeonandaeon_corePython packages, all 42 empirical BEAST phylodynamic benchmark datasets, the 50,000-generation E. coli LTEE ground-truth cohort, HIV-1 Group M Faria et al. (2014) and Korber et al. (2000) calibrations, Delphy head-to-head comparisons, 280 relaxed-clock (Miura et al. 2020) and 148,516 forward-time epidemic simulations, BV-BRC/NCBI streaming ingestion scripts,filter_bvbrc_dump_by_manifest.py, all 11 figure generation scripts,verify_manuscript_claims.py(53 deterministic ground-truth assertions),reproduce_all.sh,AGENT.md, andprompts/AGENT_REPLICATION_PLAYBOOK.md.bvbrc_influenza_a_49875_sequence_ids.tsv.gz: Exact sequence ID manifest, metadata, 72AutoClockleaf community assignments, and 453 outlier flags for the 49,875 Influenza A HA surveillance screen (no raw BV-BRC sequences are redistributed; usescripts/bvbrc_streaming/filter_bvbrc_dump_by_manifest.pyto filter future database dumps).sars2_2020_2026_clean_6317042_sequence_ids.tsv.gz(26.8 MB): Exact sequence ID manifest, release dates, collection dates, and upload lags for all 6,317,042 primary-clock SARS-CoV-2 Spike sequences streamed across 2020–2026.sars2_2020_2026_active_landmarks_1887.csv: Exact sequence IDs and coordinates for the 1,887 active online Approximate Linear Dependence (ALD) Nyström coreset landmarks (effective dimension d_eff = 1,242.2).
Notes
Files
chronaeon_nature_genetics.pdf
Files
(201.5 MB)
| Name | Size | |
|---|---|---|
|
md5:b63c251cc80a98db718d356d227f6b35
|
1.5 MB | Download |
|
md5:79fdc71a030a4439185550cc03de801b
|
1.6 MB | Preview Download |
|
md5:3c619a51eca0d289a3fd062d77210f77
|
169.8 MB | Download |
|
md5:fc3c1195fcd82b7ea91bc9b0cc43d618
|
1.6 MB | Preview Download |
|
md5:56fe115b20771edda6f1a7bf5a18dbb9
|
142.1 kB | Preview Download |
|
md5:49382d6013c7b05bf72ae473da3c5120
|
26.8 MB | Download |
Additional details
Related works
- Is supplemented by
- Project deliverable: https://veg.github.io/chronaeon/ (URL)
- Software: https://github.com/veg/HyphAeon (URL)
- Software: https://primaeon.org/time/ (URL)
Funding
- National Institutes of Health
- An in integrated platform for multiomic analyses of pathogen and host data using scalable public infrastructure 1U24AI183870-01
- National Institutes of Health
- Hypothesis Testing using Phylogenies for the 21st century 1R01GM151683-01
- U.S. National Science Foundation
- Understanding biodiversity through a global platform for assembly and analysis of large genomes 2419522