MétaCan
Menu
← Back to cohort
Record W4393409651 · doi:10.1101/2024.03.20.585782

Exact-match search with functional variant prediction enables automated DNA screening

2024· preprint· en· W4393409651 on OpenAlexaff
Dana Gretton, Brian Wang, Rey Edison, Leonard Foner, Jens Berlips, Theia Vogel, Martin Kysel, W Q Chen, Francesca Sage-Ling, Lynn Van Hauwe, Stephen Wooster, Helena Cozzarini, Benjamin Weinstein-Raun, Erika A. DeBenedictis, Andrew Bo Liu, Emma J. Chory, Hongrui Cui, Xiang Li, Jiangbin Dong, Andrés Fábrega, Otilia Don, Cassandra Tong Ye, Kaveri I. Uberoy, Ronald L. Rivest, Mingyu Gao, Yu Yu, Carsten Baum, Ivan Damgård, Andrew Chi-Chih Yao, Kevin M. Esvelt

Bibliographic record

VenuebioRxiv (Cold Spring Harbor Laboratory) · 2024
Typepreprint
Languageen
FieldBiochemistry, Genetics and Molecular Biology
TopicRNA and protein synthesis mechanisms
Canadian institutionsArtificial Intelligence in Medicine (Canada)
FundersTsinghua UniversityOpen Philanthropy ProjectAarhus Universitet
KeywordsAdversarial systemComputer scienceComputational biologyArtificial intelligenceBiology

Abstract

fetched live from OpenAlex

Abstract Custom DNA synthesis underpins modern biology, but controlled genes in the wrong hands could threaten many lives and public trust in science. In .1992, a virology-trained mass murderer tried and failed to obtain physical samples of Ebola; today, viruses can be assembled from synthetic DNA fragments. Screening orders for controlled sequences is unreliable and expensive because current similarity search algorithms yield false alarms requiring expert human review. Exact-match search can achieve perfect specificity among short known sequences by detecting subsequences unique to controlled genes, but can be trivially evaded by incorporating mutations. Here we rescue exact-match search by additionally screening for predicted functional variants of pseudo-randomly chosen subsequences that aren’t found in known unregulated genes. To experimentally assess robustness, we protected nine windows from the M13 bacteriophage virus, then invited a “red team” to launch up to 21,000 attacks at each window and measure the fitness of their designed mutants. We identified defensible windows from regulated pathogens, built a test database that our experiments indicate will block 99.999% of functional attacks, and verified its sensitivity against redesigned enzymes. Exact-match search with functional variant prediction offers a promising way to safeguard biotechnology by automating DNA synthesis screening. Summary Searching for exact matches to pre-computed functional variants unique to controlled genes enables sensitive, secure, and automated DNA synthesis screening.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.002
metaresearch head score (Gemma)0.002
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Bench or experimental · Consensus signal: Bench or experimental
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.003
Threshold uncertainty score0.011

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0020.002
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0010.001
Science and technology studies0.0000.001
Scholarly communication0.0010.001
Open science0.0010.001
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0030.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.014
GPT teacher head0.213
Teacher spread0.199 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designBench or experimental
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations3
Published2024
Admission routes1
Has abstractyes

Explore more

Same venuebioRxiv (Cold Spring Harbor Laboratory)→Same topicRNA and protein synthesis mechanisms→French-language works237,207→