A Bayesian active learning platform for scalable combination drug screens
Abstract Large-scale combination drug screens are generally considered intractable due to the immense number of possible combinations. Existing approaches use ad hoc fixed experimental designs then train machine learning models to impute unobserved combinations. Here we propose BATCHIE, an orthogona...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Nature Communications |
| Online Access: | https://doi.org/10.1038/s41467-024-55287-7 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841559271888650240 |
|---|---|
| author | Christopher Tosh Mauricio Tec Jessica B. White Jeffrey F. Quinn Glorymar Ibanez Sanchez Paul Calder Andrew L. Kung Filemon S. Dela Cruz Wesley Tansey |
| author_facet | Christopher Tosh Mauricio Tec Jessica B. White Jeffrey F. Quinn Glorymar Ibanez Sanchez Paul Calder Andrew L. Kung Filemon S. Dela Cruz Wesley Tansey |
| author_sort | Christopher Tosh |
| collection | DOAJ |
| description | Abstract Large-scale combination drug screens are generally considered intractable due to the immense number of possible combinations. Existing approaches use ad hoc fixed experimental designs then train machine learning models to impute unobserved combinations. Here we propose BATCHIE, an orthogonal approach that conducts experiments dynamically in batches. BATCHIE uses information theory and probabilistic modeling to design each batch to be maximally informative based on the results of previous experiments. On retrospective experiments from previous large-scale screens, BATCHIE designs rapidly discover highly effective and synergistic combinations. In a prospective combination screen of a library of 206 drugs on a collection of pediatric cancer cell lines, the BATCHIE model accurately predicts unseen combinations and detects synergies after exploring only 4% of the 1.4M possible experiments. Further, the model identifies a panel of top combinations for Ewing sarcomas, which follow-up validation experiments confirm to be effective, including the rational and translatable top hit of PARP plus topoisomerase I inhibition. These results demonstrate that adaptive experiments can enable large-scale unbiased combination drug screens with a relatively small number of experiments. BATCHIE is open source and publicly available ( https://github.com/tansey-lab/batchie ). |
| format | Article |
| id | doaj-art-32e6ed61b4cf4b77962e15077361cd47 |
| institution | Kabale University |
| issn | 2041-1723 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Nature Communications |
| spelling | doaj-art-32e6ed61b4cf4b77962e15077361cd472025-01-05T12:38:26ZengNature PortfolioNature Communications2041-17232025-01-0116111810.1038/s41467-024-55287-7A Bayesian active learning platform for scalable combination drug screensChristopher Tosh0Mauricio Tec1Jessica B. White2Jeffrey F. Quinn3Glorymar Ibanez Sanchez4Paul Calder5Andrew L. Kung6Filemon S. Dela Cruz7Wesley Tansey8Computational Oncology, Memorial Sloan Kettering Cancer CenterDepartment of Biostatistics, Harvard T.H. Chan School of Public HealthComputational Oncology, Memorial Sloan Kettering Cancer CenterComputational Oncology, Memorial Sloan Kettering Cancer CenterDepartment of Pediatrics, Memorial Sloan Kettering Cancer CenterDepartment of Pediatrics, Memorial Sloan Kettering Cancer CenterDepartment of Pediatrics, Memorial Sloan Kettering Cancer CenterDepartment of Pediatrics, Memorial Sloan Kettering Cancer CenterComputational Oncology, Memorial Sloan Kettering Cancer CenterAbstract Large-scale combination drug screens are generally considered intractable due to the immense number of possible combinations. Existing approaches use ad hoc fixed experimental designs then train machine learning models to impute unobserved combinations. Here we propose BATCHIE, an orthogonal approach that conducts experiments dynamically in batches. BATCHIE uses information theory and probabilistic modeling to design each batch to be maximally informative based on the results of previous experiments. On retrospective experiments from previous large-scale screens, BATCHIE designs rapidly discover highly effective and synergistic combinations. In a prospective combination screen of a library of 206 drugs on a collection of pediatric cancer cell lines, the BATCHIE model accurately predicts unseen combinations and detects synergies after exploring only 4% of the 1.4M possible experiments. Further, the model identifies a panel of top combinations for Ewing sarcomas, which follow-up validation experiments confirm to be effective, including the rational and translatable top hit of PARP plus topoisomerase I inhibition. These results demonstrate that adaptive experiments can enable large-scale unbiased combination drug screens with a relatively small number of experiments. BATCHIE is open source and publicly available ( https://github.com/tansey-lab/batchie ).https://doi.org/10.1038/s41467-024-55287-7 |
| spellingShingle | Christopher Tosh Mauricio Tec Jessica B. White Jeffrey F. Quinn Glorymar Ibanez Sanchez Paul Calder Andrew L. Kung Filemon S. Dela Cruz Wesley Tansey A Bayesian active learning platform for scalable combination drug screens Nature Communications |
| title | A Bayesian active learning platform for scalable combination drug screens |
| title_full | A Bayesian active learning platform for scalable combination drug screens |
| title_fullStr | A Bayesian active learning platform for scalable combination drug screens |
| title_full_unstemmed | A Bayesian active learning platform for scalable combination drug screens |
| title_short | A Bayesian active learning platform for scalable combination drug screens |
| title_sort | bayesian active learning platform for scalable combination drug screens |
| url | https://doi.org/10.1038/s41467-024-55287-7 |
| work_keys_str_mv | AT christophertosh abayesianactivelearningplatformforscalablecombinationdrugscreens AT mauriciotec abayesianactivelearningplatformforscalablecombinationdrugscreens AT jessicabwhite abayesianactivelearningplatformforscalablecombinationdrugscreens AT jeffreyfquinn abayesianactivelearningplatformforscalablecombinationdrugscreens AT glorymaribanezsanchez abayesianactivelearningplatformforscalablecombinationdrugscreens AT paulcalder abayesianactivelearningplatformforscalablecombinationdrugscreens AT andrewlkung abayesianactivelearningplatformforscalablecombinationdrugscreens AT filemonsdelacruz abayesianactivelearningplatformforscalablecombinationdrugscreens AT wesleytansey abayesianactivelearningplatformforscalablecombinationdrugscreens AT christophertosh bayesianactivelearningplatformforscalablecombinationdrugscreens AT mauriciotec bayesianactivelearningplatformforscalablecombinationdrugscreens AT jessicabwhite bayesianactivelearningplatformforscalablecombinationdrugscreens AT jeffreyfquinn bayesianactivelearningplatformforscalablecombinationdrugscreens AT glorymaribanezsanchez bayesianactivelearningplatformforscalablecombinationdrugscreens AT paulcalder bayesianactivelearningplatformforscalablecombinationdrugscreens AT andrewlkung bayesianactivelearningplatformforscalablecombinationdrugscreens AT filemonsdelacruz bayesianactivelearningplatformforscalablecombinationdrugscreens AT wesleytansey bayesianactivelearningplatformforscalablecombinationdrugscreens |