Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Machine Learning Algorithm for Noise Reduction and Disease-Causing Gene Feature Extraction in Gene Sequencing Data

Weichen Si, Yihao Ou, Zhen Tian

TL;DR

This work tackles the challenge of sequencing noise impairing genetic-disease analysis by introducing DeepSeqDenoise, a dual-encoder CNN–RNN noise-reduction framework, and a subsequent feature-engineering–driven model to predict pathogenic genes. The approach yields a 9.4 dB SNR improvement and a 94.3% accuracy in disease-gene prediction, validated across cardiovascular disease cohorts and independent datasets, with demonstrated clinical applicability including discovery of novel pathogenic candidates. Key contributions include a 17-feature subset for robust pathogenicity prediction, an integrated XGBoost–RF–DNN ensemble with SMOTE and Bayesian-optimized weights, and transfer learning enhancements that boost generalization to 94.8% accuracy, particularly for rare variants. Collectively, the results show strong potential for improving genetic-disease diagnosis and enabling real-time, precision medicine applications across diverse sequencing platforms.

Abstract

In this study, we propose a machine learning-based method for noise reduction and disease-causing gene feature extraction in gene sequencing DeepSeqDenoise algorithm combines CNN and RNN to effectively remove the sequencing noise, and improves the signal-to-noise ratio by 9.4 dB. We screened 17 key features by feature engineering, and constructed an integrated learning model to predict disease-causing genes with 94.3% accuracy. We successfully identified 57 new candidate disease-causing genes in a cardiovascular disease cohort validation, and detected 3 missed variants in clinical applications. The method significantly outperforms existing tools and provides strong support for accurate diagnosis of genetic diseases.

Machine Learning Algorithm for Noise Reduction and Disease-Causing Gene Feature Extraction in Gene Sequencing Data

TL;DR

This work tackles the challenge of sequencing noise impairing genetic-disease analysis by introducing DeepSeqDenoise, a dual-encoder CNN–RNN noise-reduction framework, and a subsequent feature-engineering–driven model to predict pathogenic genes. The approach yields a 9.4 dB SNR improvement and a 94.3% accuracy in disease-gene prediction, validated across cardiovascular disease cohorts and independent datasets, with demonstrated clinical applicability including discovery of novel pathogenic candidates. Key contributions include a 17-feature subset for robust pathogenicity prediction, an integrated XGBoost–RF–DNN ensemble with SMOTE and Bayesian-optimized weights, and transfer learning enhancements that boost generalization to 94.8% accuracy, particularly for rare variants. Collectively, the results show strong potential for improving genetic-disease diagnosis and enabling real-time, precision medicine applications across diverse sequencing platforms.

Abstract

In this study, we propose a machine learning-based method for noise reduction and disease-causing gene feature extraction in gene sequencing DeepSeqDenoise algorithm combines CNN and RNN to effectively remove the sequencing noise, and improves the signal-to-noise ratio by 9.4 dB. We screened 17 key features by feature engineering, and constructed an integrated learning model to predict disease-causing genes with 94.3% accuracy. We successfully identified 57 new candidate disease-causing genes in a cardiovascular disease cohort validation, and detected 3 missed variants in clinical applications. The method significantly outperforms existing tools and provides strong support for accurate diagnosis of genetic diseases.

Paper Structure

This paper contains 17 sections, 4 equations, 6 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Gene Sequencing Data Preprocessing and Noise Reduction
  3. Sequencing Data Quality Control
  4. Analysis of Data Noise Sources
  5. Machine Learning Noise Reduction Algorithm
  6. Evaluation and Comparison of Noise Reduction Effect
  7. Machine Learning-based Feature Extraction of Disease-causing Genes
  8. Identification and Screening of Gene Mutation Features
  9. Predictive Model Construction of Disease-causing Genes
  10. Gene-disease Association Analysis and Model Optimization
  11. Experimental Validation and Case Study
  12. Research Datasets and Experimental Design Scheme
  13. Experimental Results and Analysis of Noise Reduction Effect
  14. Feature Extraction and Classification Experimental Results Analysis
  15. Comparison Experiments with Existing Methods
  16. ...and 2 more sections

Figures (6)

  • Figure 1: Comparison of mass fraction distribution before and after quality control of sequencing data
  • Figure 2: Architecture of the DeepSeqDenoise algorithm
  • Figure 3: Comparison of the recovery rate of noise reduction algorithms under different noise levels
  • Figure 4: Performance of the prediction model on different types of genetic variants
  • Figure 5: Comparison of the performance of different pathogenic gene prediction methods
  • ...and 1 more figures