AI Science Tracker

Authors in Nature Communications (see paper)

2 discoveries from 2022

Discoveries

Conditional VAEs generate designer recombinases active on new target sites

Source check: Needs source check. Not independent scientific confirmation.

Finding: RecGen trained conditional variational autoencoders on >1 million Cre-like recombinase sequences evolved for 89 targets, and experimental validation showed predicted recombinases with activity on novel target sites. Abstract-level review only; success rate not read.

Technology: Conditional variational autoencoders (RecGen)

Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.

AI contribution
Generative model proposed recombinase sequences for unseen targets.
Test and result
Experimental activity assays on novel target sites.
Limitation / still unknown
Hit rate, specificity and therapeutic use not assessed here.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://europepmc.org/search?query=DOI%3A10.1038/s41467-022-35614-6

Needs source check · editorial details

The source has not been fully checked or contains conflicting information. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review scope and date are not recorded; source checked does not establish full-text review.

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 3/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

ML models co-optimize antibody affinity and specificity in novel mutational space

Source check: Needs source check. Not independent scientific confirmation.

Finding: For emibetuzumab, CDR mutant libraries were sorted for affinity and non-specific binding and deep sequenced; ML models trained on binary labels predicted continuous metrics and generalized to novel mutations, exposing trade-offs between the two properties. Abstract-level review only; truncated reading.

Technology: Machine learning models on deep-mutational sort data

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Models predicted binding properties for unseen variants to select improved antibodies.
Test and result
Predictions compared with measured binding on held-out variants.
Limitation / still unknown
Single antibody; clinical developability not assessed; full text not read.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://europepmc.org/search?query=DOI%3A10.1038/s41467-022-31457-3

Needs source check · editorial details

The source has not been fully checked or contains conflicting information. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review scope and date are not recorded; source checked does not establish full-text review.

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open the interactive profile for the timeline, or browse all organizations and model families.