A Novel Machine-Learning Approach to Predict Stress-Responsive Genes in Arabidopsis
Problem
Identifying key stress-responsive genes in Arabidopsis under salt, heat, cold, high-light and flagellin stress to enhance crop tolerance.
Why it matters
Identifying stress-tolerance genes could support development of more resilient crops.
What the AI did
Applied a hybrid ML pipeline (SMOTE, information gain, ReliefF, LASSO) with multiple classifiers to a 207-sample microarray dataset to select and classify stress-responsive genes.
Results
Identified 439 differentially expressed genes, three down-regulated; random forest outperformed other classifiers (metrics truncated in abstract).
Impact
Produced a shortlist of candidate stress-responsive genes evaluated by classifier performance.
Record
| Venue | Algorithms |
|---|---|
| Published | 2023 |
| Work type | article |
| Authors | Leyla Nazari, Vida Ghotbi, Mohammad Nadimi, Jitendra Paliwal |
| Citations | 11 (OpenAlex, at import time) |
| References | 43 |
| Model or method | Random forest (best classifier); BayesNet, logistic, MLP, SMO |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3390/a16090407 |
| Open access | gold |
| Keywords | Arabidopsis, gene selection, machine learning, SMOTE, random forest, abiotic stress |
Importance rating
real but narrower advance
Stress-tolerance gene identification could support more resilient crops, a real but early, narrow advance; no scale stated.