Accurate prediction of B-form/A-form DNA conformation propensity from primary sequence : A machine learning and free energy handshake
(2021) In Patterns 2(9).- Abstract
DNA carries the genetic code of life, with different conformations associated with different biological functions. Predicting the conformation of DNA from its primary sequence, although desirable, is a challenging problem owing to the polymorphic nature of DNA. We have deployed a host of machine learning algorithms, including the popular state-of-the-art LightGBM (a gradient boosting model), for building prediction models. We used the nested cross-validation strategy to address the issues of “overfitting” and selection bias. This simultaneously provides an unbiased estimate of the generalization performance of a machine learning algorithm and allows us to tune the hyperparameters optimally. Furthermore, we built a secondary model based... (More)
DNA carries the genetic code of life, with different conformations associated with different biological functions. Predicting the conformation of DNA from its primary sequence, although desirable, is a challenging problem owing to the polymorphic nature of DNA. We have deployed a host of machine learning algorithms, including the popular state-of-the-art LightGBM (a gradient boosting model), for building prediction models. We used the nested cross-validation strategy to address the issues of “overfitting” and selection bias. This simultaneously provides an unbiased estimate of the generalization performance of a machine learning algorithm and allows us to tune the hyperparameters optimally. Furthermore, we built a secondary model based on SHAP (SHapley Additive exPlanations) that offers crucial insight into model interpretability. Our detailed model-building strategy and robust statistical validation protocols tackle the formidable challenge of working on small datasets, which is often the case in biological and medical data.
(Less)
- author
- Gupta, Abhijit ; Kulkarni, Mandar LU and Mukherjee, Arnab
- organization
- publishing date
- 2021-09-10
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- DNA conformation, DNA sequence, DSML 2: Proof-of-Concept: Data science output has been formulated, implemented, and tested for one domain/problem, genome, LightGBM, machine learning, nested cross-validation
- in
- Patterns
- volume
- 2
- issue
- 9
- article number
- 100329
- publisher
- Cell Press
- external identifiers
-
- pmid:34553171
- scopus:85123578294
- ISSN
- 2666-3899
- DOI
- 10.1016/j.patter.2021.100329
- language
- English
- LU publication?
- yes
- id
- 8cb39788-1c38-4e4f-bc7e-76d30a77d940
- date added to LUP
- 2022-04-12 16:46:54
- date last changed
- 2025-01-14 22:08:25
@article{8cb39788-1c38-4e4f-bc7e-76d30a77d940, abstract = {{<p>DNA carries the genetic code of life, with different conformations associated with different biological functions. Predicting the conformation of DNA from its primary sequence, although desirable, is a challenging problem owing to the polymorphic nature of DNA. We have deployed a host of machine learning algorithms, including the popular state-of-the-art LightGBM (a gradient boosting model), for building prediction models. We used the nested cross-validation strategy to address the issues of “overfitting” and selection bias. This simultaneously provides an unbiased estimate of the generalization performance of a machine learning algorithm and allows us to tune the hyperparameters optimally. Furthermore, we built a secondary model based on SHAP (SHapley Additive exPlanations) that offers crucial insight into model interpretability. Our detailed model-building strategy and robust statistical validation protocols tackle the formidable challenge of working on small datasets, which is often the case in biological and medical data.</p>}}, author = {{Gupta, Abhijit and Kulkarni, Mandar and Mukherjee, Arnab}}, issn = {{2666-3899}}, keywords = {{DNA conformation; DNA sequence; DSML 2: Proof-of-Concept: Data science output has been formulated, implemented, and tested for one domain/problem; genome; LightGBM; machine learning; nested cross-validation}}, language = {{eng}}, month = {{09}}, number = {{9}}, publisher = {{Cell Press}}, series = {{Patterns}}, title = {{Accurate prediction of B-form/A-form DNA conformation propensity from primary sequence : A machine learning and free energy handshake}}, url = {{http://dx.doi.org/10.1016/j.patter.2021.100329}}, doi = {{10.1016/j.patter.2021.100329}}, volume = {{2}}, year = {{2021}}, }