Pon-tstab : Protein variant stability predictor. importance of training data quality
(2018) In International Journal of Molecular Sciences 19(4).- Abstract
Several methods have been developed to predict effects of amino acid substitutions on protein stability. Benchmark datasets are essential for method training and testing and have numerous requirements including that the data is representative for the investigated phenomenon. Available machine learning algorithms for variant stability have all been trained with ProTherm data. We noticed a number of issues with the contents, quality and relevance of the database. There were errors, but also features that had not been clearly communicated. Consequently, all machine learning variant stability predictors have been trained on biased and incorrect data. We obtained a corrected dataset and trained a random forests-based tool, PON-tstab,... (More)
Several methods have been developed to predict effects of amino acid substitutions on protein stability. Benchmark datasets are essential for method training and testing and have numerous requirements including that the data is representative for the investigated phenomenon. Available machine learning algorithms for variant stability have all been trained with ProTherm data. We noticed a number of issues with the contents, quality and relevance of the database. There were errors, but also features that had not been clearly communicated. Consequently, all machine learning variant stability predictors have been trained on biased and incorrect data. We obtained a corrected dataset and trained a random forests-based tool, PON-tstab, applicable to variants in any organism. Our results highlight the importance of the benchmark quality, suitability and appropriateness. Predictions are provided for three categories: stability decreasing, increasing and those not affecting stability.
(Less)
- author
- Yang, Yang LU ; Urolagin, Siddhaling LU ; Niroula, Abhishek LU ; Ding, Xuesong ; Shen, Bairong and Vihinen, Mauno LU
- organization
- publishing date
- 2018-04-01
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Benchmark quality, Machine learning method, Mutation, Protein stability prediction, Variation interpretation
- in
- International Journal of Molecular Sciences
- volume
- 19
- issue
- 4
- article number
- 1009
- publisher
- MDPI AG
- external identifiers
-
- scopus:85044635395
- pmid:29597263
- ISSN
- 1661-6596
- DOI
- 10.3390/ijms19041009
- language
- English
- LU publication?
- yes
- id
- 19e2a293-574a-4411-924f-b46f42b39d98
- date added to LUP
- 2018-04-10 12:38:14
- date last changed
- 2024-08-06 16:19:24
@article{19e2a293-574a-4411-924f-b46f42b39d98, abstract = {{<p>Several methods have been developed to predict effects of amino acid substitutions on protein stability. Benchmark datasets are essential for method training and testing and have numerous requirements including that the data is representative for the investigated phenomenon. Available machine learning algorithms for variant stability have all been trained with ProTherm data. We noticed a number of issues with the contents, quality and relevance of the database. There were errors, but also features that had not been clearly communicated. Consequently, all machine learning variant stability predictors have been trained on biased and incorrect data. We obtained a corrected dataset and trained a random forests-based tool, PON-tstab, applicable to variants in any organism. Our results highlight the importance of the benchmark quality, suitability and appropriateness. Predictions are provided for three categories: stability decreasing, increasing and those not affecting stability.</p>}}, author = {{Yang, Yang and Urolagin, Siddhaling and Niroula, Abhishek and Ding, Xuesong and Shen, Bairong and Vihinen, Mauno}}, issn = {{1661-6596}}, keywords = {{Benchmark quality; Machine learning method; Mutation; Protein stability prediction; Variation interpretation}}, language = {{eng}}, month = {{04}}, number = {{4}}, publisher = {{MDPI AG}}, series = {{International Journal of Molecular Sciences}}, title = {{Pon-tstab : Protein variant stability predictor. importance of training data quality}}, url = {{http://dx.doi.org/10.3390/ijms19041009}}, doi = {{10.3390/ijms19041009}}, volume = {{19}}, year = {{2018}}, }