ExhauFS: exhaustive search-based feature selection for classification and survival regression

Abstract

Feature selection is one of the main techniques used to prevent overfitting in machine learning applications. The most straightforward approach for feature selection is an exhaustive search: one can go over all possible feature combinations and pick up the model with the highest accuracy. This method together with its optimizations were actively used in biomedical research, however, publicly available implementation is missing. We present ExhauFS—the user-friendly command-line implementation of the exhaustive search approach for classification and survival regression. Aside from tool description, we included three application examples in the manuscript to comprehensively review the implemented functionality. First, we executed ExhauFS on a toy cervical cancer dataset to illustrate basic concepts. Then, multi-cohort microarray breast cancer datasets were used to construct gene signatures for 5-year recurrence classification. The vast majority of signatures constructed by ExhauFS passed 0.65 threshold of sensitivity and specificity on all datasets, including the validation one. Moreover, a number of gene signatures demonstrated reliable performance on independent RNA-seq dataset without any coefficient re-tuning, i.e., turned out to be cross-platform. Finally, Cox survival regression models were used to fit isomiR signatures for overall survival prediction for patients with colorectal cancer. Similarly to the previous example, the major part of models passed the pre-defined concordance index threshold 0.65 on all datasets. In both real-world scenarios (breast and colorectal cancer datasets), ExhauFS was benchmarked against state-of-the-art feature selection models, including L1-regularized sparse models. In case of breast cancer, we were unable to construct reliable cross-platform classifiers using alternative feature selection approaches. In case of colorectal cancer not a single model passed the same 0.65 threshold. Source codes and documentation of ExhauFS are available on GitHub: https://github.com/s-a-nersisyan/ExhauFS.

A novel approach for a joint analysis of isomiR and mRNA expression data reveals features of isomiR targeting in breast cancer

“…Since 3′-end isomiR variations do not alter the isomiR targetome much (seed region not affected), we summed up the expression of isomiRs originating from the same miRNA, which have identical 5′-end sequences (5′-isomiRs). We used our previous notation for 5′-isomiRs ( Nersisyan et al, 2022b ): a number after the “|” symbol stands for a shift from the canonical 5′-end in the 5′–3′ direction. For example, hsa-miR-192-5p|+1 differs from the canonical hsa-miR-192-5p miRNA by the absence of the first nucleotide on its 5′-end.…”

Circulating miRNA profiles in COVID-19 patients and meta-analysis: implications for disease progression and prognosis

“…Therefore, we propose that building machine learning models based on meta-analysis of multiple studies would enhance the robustness and reliability of these models, especially given the increasing availability of data and recent publications in this field. In addition, computational tools such as ExhauFS were employed to identify top ranked features for machine learning predictions 16 . For model development and validation, our study (n = 18, 27 for disease, and n = 77, 27 for severity) served as the training set, while the study by Zeng et al (n = 61, 48 for disease and n = 52, 48 for severity) was utilized as the filtration set, with the Guttman Study (n = 12, 18 for disease and n = 18, 18 for severity) and Garcia Study served as validation sets (n = 13, 15 for severity, with no healthy controls).…”

“…Notably, we observed a significant down-regulation of a large cluster of miRNAs located on chromosome 14 (14q32), comprising over 90 members. We built machine learning models based on meta-analysis results combined with an exhaustive feature selection tool 16 that was able to predict COVID-19 disease and severity across multiple independently published studies. Our results suggest a robust method for building miRNA-based models for disease diagnosis and prognosis and highlight overlapping roles of different miRNA biomarkers.…”

Enhanced Weak Spatial Modeling Through CNN-Based Deep Sign Language Skeletal Feature Transformation

“…Exhaustive feature searching is a brute-force technique that evaluates all possible subsets of features and selects the one that maximizes a given criterion function [56], [57]. Let's assume that a set of…”