{"publication":{"abstract":"Attention mechanisms are increasingly applied to genotype–phenotype mapping problems, particularly for capturing epistatic interactions. Rijal et al. (2025) recently demonstrated an attention-based model for this task, but their architecture omitted standard transformer components like skip connections, layer normalization, and feed-forward sub-layers.\n\nHere, we test whether incorporating these canonical elements improves predictive performance. Using the same yeast dataset (~100,000 segregants, 18 growth phenotypes), we show that standard transformer components moderately improve accuracy. We also find that predicting all phenotypes jointly provides additional gains by leveraging cross-phenotype genetic correlations, an advantage the original single-output approach couldn't exploit.\n\nThis work should interest researchers applying deep learning to genotype–phenotype problems. Our results suggest that well-established architectural choices from the broader ML literature transfer well to genetics applications, and that multi-task learning offers a straightforward path to improved predictions when correlated phenotypes are available. We share all code and model checkpoints to enable rapid iteration by others.","body":"# View the notebook\n\n::::::only{platform=web}\n:::::div{.info-box}\nThe **full pub** is available [here](https://arcadia-science.github.io/2025-geno-pheno-attention/).\n\nThe **source code** to generate it is available in [this GitHub repo](https://github.com/Arcadia-Science/2025-geno-pheno-attention) (DOI\\: [10.5281/zenodo.15320438](https://doi.org/10.5281/zenodo.15320438)).\n:::::\n::::::\n\nIn the future, we hope to host [notebook pubs](https://dx.doi.org/10.57844/arcadia-ca21-23bb) directly on PubPub. Until that’s possible, we’ll create stubs like this with key metadata like the DOI, author roles, citation information, and an external link to the pub itself.","contributors":[{"user_id":4218,"role":"Supervision","first_name":"Brae M.","last_name":"Bigge","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Brae-Arcadia-headshot.png","in_byline":1,"priority":1,"affiliations":[]},{"user_id":4229,"role":"Conceptualization","first_name":"Evan","last_name":"Kiefl","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Evan-Arcadia-headshot.png","in_byline":1,"priority":2,"affiliations":[]},{"user_id":4229,"role":"Formal Analysis","first_name":"Evan","last_name":"Kiefl","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Evan-Arcadia-headshot.png","in_byline":1,"priority":2,"affiliations":[]},{"user_id":4229,"role":"Investigation","first_name":"Evan","last_name":"Kiefl","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Evan-Arcadia-headshot.png","in_byline":1,"priority":2,"affiliations":[]},{"user_id":4229,"role":"Software","first_name":"Evan","last_name":"Kiefl","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Evan-Arcadia-headshot.png","in_byline":1,"priority":2,"affiliations":[]},{"user_id":4229,"role":"Validation","first_name":"Evan","last_name":"Kiefl","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Evan-Arcadia-headshot.png","in_byline":0,"priority":2,"affiliations":[]},{"user_id":4229,"role":"Writing","first_name":"Evan","last_name":"Kiefl","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Evan-Arcadia-headshot.png","in_byline":1,"priority":2,"affiliations":[]},{"user_id":4225,"role":"Conceptualization","first_name":"Erin","last_name":"McGeever","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Erin-Arcadia-headshot.png","in_byline":1,"priority":3,"affiliations":[]},{"user_id":4225,"role":"Data Curation","first_name":"Erin","last_name":"McGeever","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Erin-Arcadia-headshot.png","in_byline":0,"priority":3,"affiliations":[]},{"user_id":4225,"role":"Investigation","first_name":"Erin","last_name":"McGeever","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Erin-Arcadia-headshot.png","in_byline":1,"priority":3,"affiliations":[]},{"user_id":4225,"role":"Software","first_name":"Erin","last_name":"McGeever","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Erin-Arcadia-headshot.png","in_byline":1,"priority":3,"affiliations":[]},{"user_id":4225,"role":"Validation","first_name":"Erin","last_name":"McGeever","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Erin-Arcadia-headshot.png","in_byline":0,"priority":3,"affiliations":[]},{"user_id":4230,"role":"Editing","first_name":"George","last_name":"Sandler","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/George-Arcadia-headshot.png","in_byline":0,"priority":4,"affiliations":[]},{"user_id":4222,"role":"Supervision","first_name":"Ryan","last_name":"York","avatar":"https://thestacks-01.s3.amazonaws.com/users/avatars/Ryan-Arcadia-headshot.png","in_byline":1,"priority":5,"affiliations":[]}],"created_at":"2025-05-01T18:30:00.000Z","doi":"10.57844/arcadia-bmb9-fzxd","version_doi":null,"feedback_form_embed_src":null,"id":25,"license":"CC BY","linked_assets":[{"type":"Code","url":"https://github.com/Arcadia-Science/2025-geno-pheno-attention","name":"2025-geno-pheno-attention"}],"orgs":[{"org_id":2,"priority":null,"slug":"arcadia-science","name":"Arcadia Science","avatar":"avatars/orgs/org_2_1766013565.jpg"}],"pdf_url":"https://thestacks-01.s3.us-west-2.amazonaws.com/publications/observation-geno-pheno-attention/observation-geno-pheno-attention_v2.pdf","references":[],"slug":"observation-geno-pheno-attention","social_posts_count":null,"social_posts_embed_src":"https://publishing-tools.arcadiascience.com/twitter/870067101194838913","state":"PUBLISHED","subtitle":"We added standard transformer components, omitted by Rijal et al. (2025) in their attention-based genotype–phenotype mapping. We found that this addition substantially boosts predictive accuracy on their yeast dataset.","tags":["Observation","not actively updating"],"title":"Cross-trait learning with a canonical transformer tops custom attention in genotype–phenotype mapping","version_desc":"Added \"Purpose\" section to stub","version_number":2,"versions":[{"id":370,"version_number":2,"version_desc":"Added \"Purpose\" section to stub","doi":null,"created_at":"2025-12-12T22:12:02.000Z"},{"id":97,"version_number":1,"version_desc":null,"doi":null,"created_at":"2025-05-01T18:30:00.000Z"}]}}