Machine Learning QTL Mapping Research

Deep Learning Model Development for QTL Detection

Interns will develop and optimize convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to identify quantitative trait loci from genomic and phenotypic datasets. They will work with frameworks like TensorFlow and PyTorch to build models that outperform traditional statistical methods in detecting complex trait associations.

Genomic Data Preprocessing and Feature Engineering

Interns will process large-scale genomic datasets including SNP arrays and whole-genome sequencing data, performing quality control, imputation, and feature selection. They will develop pipelines to extract meaningful features from raw genetic data for use in machine learning QTL mapping algorithms.

Phenotype Prediction Using Genomic Selection Models

Interns will build and validate machine learning models that predict quantitative traits from genomic markers using algorithms such as random forests, gradient boosting, and support vector machines. They will evaluate model performance through cross-validation and contribute to genomic selection breeding programs.

Multi-trait QTL Analysis and Network Integration

Interns will apply machine learning techniques to identify pleiotropic QTLs affecting multiple traits simultaneously and integrate gene networks with QTL mapping results. They will work with tools for pathway analysis and systems genetics to understand genetic architecture of complex traits.

Population Structure Correction and Association Study Optimization

Interns will implement machine learning methods to correct for population stratification and kinship structure in genome-wide association studies (GWAS) for improved QTL mapping accuracy. They will develop and optimize algorithms to increase statistical power while reducing false positives in trait-marker associations.