Predicting Cancer Drug Response In Vivo by Learning an Optimal Feature Selection of Tumour Molecular Profiles
(1) Background: Inter-tumour heterogeneity is one of cancer’s most fundamental features. Patient stratification based on drug response prediction is hence needed for effective anti-cancer therapy. However, single-gene markers of response are rare and/or may fail to achieve a significant impact in th...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-09-01
|
| Series: | Biomedicines |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-9059/9/10/1319 |
| _version_ | 1827680120954945536 |
|---|---|
| author | Linh C. Nguyen Stefan Naulaerts Alejandra Bruna Ghita Ghislat Pedro J. Ballester |
| author_facet | Linh C. Nguyen Stefan Naulaerts Alejandra Bruna Ghita Ghislat Pedro J. Ballester |
| author_sort | Linh C. Nguyen |
| collection | DOAJ |
| description | (1) Background: Inter-tumour heterogeneity is one of cancer’s most fundamental features. Patient stratification based on drug response prediction is hence needed for effective anti-cancer therapy. However, single-gene markers of response are rare and/or may fail to achieve a significant impact in the clinic. Machine Learning (ML) is emerging as a particularly promising complementary approach to precision oncology. (2) Methods: Here we leverage comprehensive Patient-Derived Xenograft (PDX) pharmacogenomic data sets with dimensionality-reducing ML algorithms with this purpose. (3) Results: Combining multiple gene alterations via ML leads to better discrimination between sensitive and resistant PDXs in 19 of the 26 analysed cases. Highly predictive ML models employing concise gene lists were found for three cases: paclitaxel (breast cancer), binimetinib (breast cancer) and cetuximab (colorectal cancer). Interestingly, each of these multi-gene ML models identifies some treatment-responsive PDXs not harbouring the best actionable mutation for that case. Thus, ML multi-gene predictors generally have much fewer false negatives than the corresponding single-gene marker. (4) Conclusions: As PDXs often recapitulate clinical outcomes, these results suggest that many more patients could benefit from precision oncology if ML algorithms were also applied to existing clinical pharmacogenomics data, especially those algorithms generating classifiers combining data-selected gene alterations. |
| first_indexed | 2024-03-10T06:43:32Z |
| format | Article |
| id | doaj.art-865f23335498479d9afd376cc1ca8346 |
| institution | Directory Open Access Journal |
| issn | 2227-9059 |
| language | English |
| last_indexed | 2024-03-10T06:43:32Z |
| publishDate | 2021-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Biomedicines |
| spelling | doaj.art-865f23335498479d9afd376cc1ca83462023-11-22T17:30:00ZengMDPI AGBiomedicines2227-90592021-09-01910131910.3390/biomedicines9101319Predicting Cancer Drug Response In Vivo by Learning an Optimal Feature Selection of Tumour Molecular ProfilesLinh C. Nguyen0Stefan Naulaerts1Alejandra Bruna2Ghita Ghislat3Pedro J. Ballester4Cancer Research Center of Marseille, INSERM U1068, F-13009 Marseille, FranceLudwig Institute for Cancer Research, 1200 Brussels, BelgiumThe Institute of Cancer Research, London SM2 5NG, UKCentre d’Immunologie de Marseille-Luminy, INSERM U1104, CNRS UMR7280, F-13009 Marseille, FranceCancer Research Center of Marseille, INSERM U1068, F-13009 Marseille, France(1) Background: Inter-tumour heterogeneity is one of cancer’s most fundamental features. Patient stratification based on drug response prediction is hence needed for effective anti-cancer therapy. However, single-gene markers of response are rare and/or may fail to achieve a significant impact in the clinic. Machine Learning (ML) is emerging as a particularly promising complementary approach to precision oncology. (2) Methods: Here we leverage comprehensive Patient-Derived Xenograft (PDX) pharmacogenomic data sets with dimensionality-reducing ML algorithms with this purpose. (3) Results: Combining multiple gene alterations via ML leads to better discrimination between sensitive and resistant PDXs in 19 of the 26 analysed cases. Highly predictive ML models employing concise gene lists were found for three cases: paclitaxel (breast cancer), binimetinib (breast cancer) and cetuximab (colorectal cancer). Interestingly, each of these multi-gene ML models identifies some treatment-responsive PDXs not harbouring the best actionable mutation for that case. Thus, ML multi-gene predictors generally have much fewer false negatives than the corresponding single-gene marker. (4) Conclusions: As PDXs often recapitulate clinical outcomes, these results suggest that many more patients could benefit from precision oncology if ML algorithms were also applied to existing clinical pharmacogenomics data, especially those algorithms generating classifiers combining data-selected gene alterations.https://www.mdpi.com/2227-9059/9/10/1319biomarker discoverymachine learningpatient-derived xenograftprecision oncologytumour profiling |
| spellingShingle | Linh C. Nguyen Stefan Naulaerts Alejandra Bruna Ghita Ghislat Pedro J. Ballester Predicting Cancer Drug Response In Vivo by Learning an Optimal Feature Selection of Tumour Molecular Profiles Biomedicines biomarker discovery machine learning patient-derived xenograft precision oncology tumour profiling |
| title | Predicting Cancer Drug Response In Vivo by Learning an Optimal Feature Selection of Tumour Molecular Profiles |
| title_full | Predicting Cancer Drug Response In Vivo by Learning an Optimal Feature Selection of Tumour Molecular Profiles |
| title_fullStr | Predicting Cancer Drug Response In Vivo by Learning an Optimal Feature Selection of Tumour Molecular Profiles |
| title_full_unstemmed | Predicting Cancer Drug Response In Vivo by Learning an Optimal Feature Selection of Tumour Molecular Profiles |
| title_short | Predicting Cancer Drug Response In Vivo by Learning an Optimal Feature Selection of Tumour Molecular Profiles |
| title_sort | predicting cancer drug response in vivo by learning an optimal feature selection of tumour molecular profiles |
| topic | biomarker discovery machine learning patient-derived xenograft precision oncology tumour profiling |
| url | https://www.mdpi.com/2227-9059/9/10/1319 |
| work_keys_str_mv | AT linhcnguyen predictingcancerdrugresponseinvivobylearninganoptimalfeatureselectionoftumourmolecularprofiles AT stefannaulaerts predictingcancerdrugresponseinvivobylearninganoptimalfeatureselectionoftumourmolecularprofiles AT alejandrabruna predictingcancerdrugresponseinvivobylearninganoptimalfeatureselectionoftumourmolecularprofiles AT ghitaghislat predictingcancerdrugresponseinvivobylearninganoptimalfeatureselectionoftumourmolecularprofiles AT pedrojballester predictingcancerdrugresponseinvivobylearninganoptimalfeatureselectionoftumourmolecularprofiles |