Doppelgänger spotting in biomedical gene expression data
Summary: Doppelgänger effects (DEs) occur when samples exhibit chance similarities such that, when split across training and validation sets, inflates the trained machine learning (ML) model performance. This inflationary effect causes misleading confidence on the deployability of the model. Thus, s...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2022-08-01
|
| Series: | iScience |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2589004222010604 |
| _version_ | 1818498642406801408 |
|---|---|
| author | Li Rong Wang Xin Yun Choy Wilson Wen Bin Goh |
| author_facet | Li Rong Wang Xin Yun Choy Wilson Wen Bin Goh |
| author_sort | Li Rong Wang |
| collection | DOAJ |
| description | Summary: Doppelgänger effects (DEs) occur when samples exhibit chance similarities such that, when split across training and validation sets, inflates the trained machine learning (ML) model performance. This inflationary effect causes misleading confidence on the deployability of the model. Thus, so far, there are no tools for doppelgänger identification or standard practices to manage their confounding implications. We present doppelgangerIdentifier, a software suite for doppelgänger identification and verification. Applying doppelgangerIdentifier across a multitude of diseases and data types, we show the pervasive nature of DEs in biomedical gene expression data. We also provide guidelines toward proper doppelgänger identification by exploring the ramifications of lingering batch effects from batch imbalances on the sensitivity of our doppelgänger identification algorithm. We suggest doppelgänger verification as a useful procedure to establish baselines for model evaluation that may inform on whether feature selection and ML on the data set may yield meaningful insights. |
| first_indexed | 2024-12-10T20:18:21Z |
| format | Article |
| id | doaj.art-5d1e1ca92eb7484f86a83b7332274446 |
| institution | Directory Open Access Journal |
| issn | 2589-0042 |
| language | English |
| last_indexed | 2024-12-10T20:18:21Z |
| publishDate | 2022-08-01 |
| publisher | Elsevier |
| record_format | Article |
| series | iScience |
| spelling | doaj.art-5d1e1ca92eb7484f86a83b73322744462022-12-22T01:35:07ZengElsevieriScience2589-00422022-08-01258104788Doppelgänger spotting in biomedical gene expression dataLi Rong Wang0Xin Yun Choy1Wilson Wen Bin Goh2School of Computer Science and Engineering, Nanyang Technological University, 60 Nanyang Drive, 637551, SingaporeSchool of Computer Science and Engineering, Nanyang Technological University, 60 Nanyang Drive, 637551, SingaporeSchool of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, 637551, Singapore; Lee Kong Chian School of Medicine, Nanyang Technological University, 60 Nanyang Drive, 637551, Singapore; Centre for Biomedical Informatics, Nanyang Technological University, 60 Nanyang Drive, 637551, Singapore; Corresponding authorSummary: Doppelgänger effects (DEs) occur when samples exhibit chance similarities such that, when split across training and validation sets, inflates the trained machine learning (ML) model performance. This inflationary effect causes misleading confidence on the deployability of the model. Thus, so far, there are no tools for doppelgänger identification or standard practices to manage their confounding implications. We present doppelgangerIdentifier, a software suite for doppelgänger identification and verification. Applying doppelgangerIdentifier across a multitude of diseases and data types, we show the pervasive nature of DEs in biomedical gene expression data. We also provide guidelines toward proper doppelgänger identification by exploring the ramifications of lingering batch effects from batch imbalances on the sensitivity of our doppelgänger identification algorithm. We suggest doppelgänger verification as a useful procedure to establish baselines for model evaluation that may inform on whether feature selection and ML on the data set may yield meaningful insights.http://www.sciencedirect.com/science/article/pii/S2589004222010604BioinformaticsGenomicsHuman Genetics |
| spellingShingle | Li Rong Wang Xin Yun Choy Wilson Wen Bin Goh Doppelgänger spotting in biomedical gene expression data iScience Bioinformatics Genomics Human Genetics |
| title | Doppelgänger spotting in biomedical gene expression data |
| title_full | Doppelgänger spotting in biomedical gene expression data |
| title_fullStr | Doppelgänger spotting in biomedical gene expression data |
| title_full_unstemmed | Doppelgänger spotting in biomedical gene expression data |
| title_short | Doppelgänger spotting in biomedical gene expression data |
| title_sort | doppelganger spotting in biomedical gene expression data |
| topic | Bioinformatics Genomics Human Genetics |
| url | http://www.sciencedirect.com/science/article/pii/S2589004222010604 |
| work_keys_str_mv | AT lirongwang doppelgangerspottinginbiomedicalgeneexpressiondata AT xinyunchoy doppelgangerspottinginbiomedicalgeneexpressiondata AT wilsonwenbingoh doppelgangerspottinginbiomedicalgeneexpressiondata |