Generating experimentally unrelated target molecule-binding highly functionalized nucleic-acid polymers using machine learning
In vitro library screening is a powerful approach to identify functional biopolymers, but only covers a fraction of possible sequences. Here, the authors use experimental in vitro selection results to train a conditional variational autoencoder machine learning model that generates biopolymers with...
Main Authors: | , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2022-08-01
|
Series: | Nature Communications |
Online Access: | https://doi.org/10.1038/s41467-022-31955-4 |
_version_ | 1811343872959184896 |
---|---|
author | Jonathan C. Chen Jonathan P. Chen Max W. Shen Michael Wornow Minwoo Bae Wei-Hsi Yeh Alvin Hsu David R. Liu |
author_facet | Jonathan C. Chen Jonathan P. Chen Max W. Shen Michael Wornow Minwoo Bae Wei-Hsi Yeh Alvin Hsu David R. Liu |
author_sort | Jonathan C. Chen |
collection | DOAJ |
description | In vitro library screening is a powerful approach to identify functional biopolymers, but only covers a fraction of possible sequences. Here, the authors use experimental in vitro selection results to train a conditional variational autoencoder machine learning model that generates biopolymers with no apparent sequence similarity to experimentally derived examples, but that nevertheless bind the target molecule with similar potent binding affinity. |
first_indexed | 2024-04-13T19:36:30Z |
format | Article |
id | doaj.art-502e0dcaf2ca4731b191cf430212bb7e |
institution | Directory Open Access Journal |
issn | 2041-1723 |
language | English |
last_indexed | 2024-04-13T19:36:30Z |
publishDate | 2022-08-01 |
publisher | Nature Portfolio |
record_format | Article |
series | Nature Communications |
spelling | doaj.art-502e0dcaf2ca4731b191cf430212bb7e2022-12-22T02:33:00ZengNature PortfolioNature Communications2041-17232022-08-0113111710.1038/s41467-022-31955-4Generating experimentally unrelated target molecule-binding highly functionalized nucleic-acid polymers using machine learningJonathan C. Chen0Jonathan P. Chen1Max W. Shen2Michael Wornow3Minwoo Bae4Wei-Hsi Yeh5Alvin Hsu6David R. Liu7Merkin Institute of Transformative Technologies in Healthcare, Broad Institute of Harvard and MITWork conducted at Uber AI Labs, Uber Technologies, Inc.Merkin Institute of Transformative Technologies in Healthcare, Broad Institute of Harvard and MITMerkin Institute of Transformative Technologies in Healthcare, Broad Institute of Harvard and MITMerkin Institute of Transformative Technologies in Healthcare, Broad Institute of Harvard and MITMerkin Institute of Transformative Technologies in Healthcare, Broad Institute of Harvard and MITMerkin Institute of Transformative Technologies in Healthcare, Broad Institute of Harvard and MITMerkin Institute of Transformative Technologies in Healthcare, Broad Institute of Harvard and MITIn vitro library screening is a powerful approach to identify functional biopolymers, but only covers a fraction of possible sequences. Here, the authors use experimental in vitro selection results to train a conditional variational autoencoder machine learning model that generates biopolymers with no apparent sequence similarity to experimentally derived examples, but that nevertheless bind the target molecule with similar potent binding affinity.https://doi.org/10.1038/s41467-022-31955-4 |
spellingShingle | Jonathan C. Chen Jonathan P. Chen Max W. Shen Michael Wornow Minwoo Bae Wei-Hsi Yeh Alvin Hsu David R. Liu Generating experimentally unrelated target molecule-binding highly functionalized nucleic-acid polymers using machine learning Nature Communications |
title | Generating experimentally unrelated target molecule-binding highly functionalized nucleic-acid polymers using machine learning |
title_full | Generating experimentally unrelated target molecule-binding highly functionalized nucleic-acid polymers using machine learning |
title_fullStr | Generating experimentally unrelated target molecule-binding highly functionalized nucleic-acid polymers using machine learning |
title_full_unstemmed | Generating experimentally unrelated target molecule-binding highly functionalized nucleic-acid polymers using machine learning |
title_short | Generating experimentally unrelated target molecule-binding highly functionalized nucleic-acid polymers using machine learning |
title_sort | generating experimentally unrelated target molecule binding highly functionalized nucleic acid polymers using machine learning |
url | https://doi.org/10.1038/s41467-022-31955-4 |
work_keys_str_mv | AT jonathancchen generatingexperimentallyunrelatedtargetmoleculebindinghighlyfunctionalizednucleicacidpolymersusingmachinelearning AT jonathanpchen generatingexperimentallyunrelatedtargetmoleculebindinghighlyfunctionalizednucleicacidpolymersusingmachinelearning AT maxwshen generatingexperimentallyunrelatedtargetmoleculebindinghighlyfunctionalizednucleicacidpolymersusingmachinelearning AT michaelwornow generatingexperimentallyunrelatedtargetmoleculebindinghighlyfunctionalizednucleicacidpolymersusingmachinelearning AT minwoobae generatingexperimentallyunrelatedtargetmoleculebindinghighlyfunctionalizednucleicacidpolymersusingmachinelearning AT weihsiyeh generatingexperimentallyunrelatedtargetmoleculebindinghighlyfunctionalizednucleicacidpolymersusingmachinelearning AT alvinhsu generatingexperimentallyunrelatedtargetmoleculebindinghighlyfunctionalizednucleicacidpolymersusingmachinelearning AT davidrliu generatingexperimentallyunrelatedtargetmoleculebindinghighlyfunctionalizednucleicacidpolymersusingmachinelearning |