#32 · ML / AI methods
ML-guided design of artificial microRNAs and functional consequences of miRNA gene…
The dominant thread is computational handling of small RNA sequence and processing. miRarchitect, a preprint, is a web-based ML platform that integrates target-site selection, siRNA insert design, scaffold choice and off-target assessment for artificial microRNAs. FragmentFinder, a published Windows-executable tool, identifies and characterizes short RNA fragments excised from any sncRNA across species. A published TCGA analysis of more than 10,000 cancers reports that many somatic miRNA gene mutations are deleterious, altering mature miRNA levels and isomiR profiles. Together these items connect variation in miRNA and sncRNA products to tool-level design and annotation. The available summaries do not establish experimental validation of the designed constructs, and FragmentFinder is not described as machine learning-based.
Preprints
TOXsiRNA: A web server to predict the toxicity of chemically modified siRNAs
TOXsiRNA web server predicts toxicity of chemically modified siRNAs using SVM, LR, KNN and ANN models; best SVM model reached PCC 0.92 on independent…
Machine learning-guided design of artificial microRNAs for targeted gene silencing
miRarchitect, a web-based ML platform for designing artificial microRNAs, integrates target-site selection, siRNA insert design, scaffold choice and off-target…
RNAiSpline: A Deep learning model for siRNA efficacy prediction
RNAiSpline: a deep learning model for siRNA efficacy prediction using self-supervised pretraining with KAN, CNN, and Transformer encoder; reports ROC-AUC…
Topic history
| Date | Rank |
|---|---|
| #32 | |
| #40 | |
| #34 |