Multimodal Evidence-based Locus Prioritization (MELP) Algorithm
MELP is a method designed for efficient mining of key genes underlying target traits by leveraging population-scale multi‑omics data and enhancing gene signals through multi‑level detection.
The MELP pipeline consists of two main steps:
Step 1: Signal detection across individual omics layers using conventional approaches
Examples include GWAS, TWAS, eQTL analysis, etc. The analysis codes used in this project are provided below:
• GWAS (MLM model in TASSEL):
sh 1_gwas.sh
• TWAS (rrBLUP method):
Rscript 2_twas.r
• eQTL analysis (EMMAX software):
sh 3_emmax.sh
• Co‑expression analysis:
Rscript 4_coexpress.r
Step 2: Integration of multi‑omics signals to obtain association scores for all genes
Run the integration script:
python3 5_integration_features.py