Yomics

Analysis and figure-generation code for a multi-omic study of the yeast translational response to acute glucose starvation (Saccharomyces cerevisiae), comparing non-starved (NS) and 10-minute starved (S10) conditions.

The work extends the eTCP-seq / Stochastic Translation Efficiency (STE) framework of Horvath et al. (2024, Nucleic Acids Research) with direct RNA sequencing (DRS) of polysome-fractionated transcripts, jointly measuring transcript abundance, degradation, poly(A) tail length, and three RNA modifications (m5C, m6A, pseudouridine) from native molecules.

A lasso-penalized mixture-of-regressions model (flexmix + glmnet, k=5) fit on 29 candidate features across 5,033 transcripts identifies five regulatory clusters with distinct translational responses to starvation.

Repository structure

• PreProcessing/ — shell pipelines for degradation quantification (INDEGRA), RNA modification calling, and poly(A) tail length estimation from raw sequencing data.
• Entropy/ — k-mer entropy and nucleotide composition analyses of UTRs and CDS.
• Multi_Omic_Integration/ — R scripts building the integrated per-transcript feature table, fitting the lasso mixture-of-regressions model, extracting cluster assignments, feature importances (SHAP), and GO enrichment.
• Figures/ — one folder per main and supplementary figure, containing the R (and where relevant, Python) scripts used to generate it from the processed data tables.
• common/ — shared Python helper module (_common.py) used by the cis-regulatory feature scripts under Figures/.

Requirements

R (≥ 4.x): ggplot2, patchwork, cowplot, pheatmap, flexmix, glmnet, DESeq2, clusterProfiler, svglite, scales, caret, dplyr, tibble, purrr

Python (≥ 3.9): pandas, numpy, scipy, requests, matplotlib

Environment variables

Scripts read data locations from environment variables rather than hardcoded paths. Set these before running anything:

Variable	Description
YEAST_PROJECT_DIR	Root of the analysis project (contains Analysis/, Paper/, etc.)
YEAST_DATA_DIR	Root of the primary sequencing-derived data (Annotation/, Final_db/, Structure/, etc.)
YEAST_SUMMARY_TABLE	Path to the combined per-transcript summary CSV
YEAST_POLYA_RAW_DIR	Path to raw poly(A) tail length output files
YEAST_REGULON_DIR	Path to the Regulon Extension pipeline outputs (cis-regulatory feature tables)
YEAST_CLUSTER_DIR	Path to the per-cluster gene list files (genes_cluster_lasso_M1_k5_cl*.txt)
YEAST_FIGURE_OUT	Output directory for rendered figures

Example:

export YEAST_PROJECT_DIR=/path/to/NewMapping
export YEAST_DATA_DIR=/path/to/Yeast
export YEAST_SUMMARY_TABLE=/path/to/combine_table_v21.csv
export YEAST_POLYA_RAW_DIR="/path/to/PolyA/Raw File"
export YEAST_REGULON_DIR=/path/to/Regulon_pipeline
export YEAST_CLUSTER_DIR=/path/to/Clusters
export YEAST_FIGURE_OUT=/path/to/output/figures

Reproducibility

Random seeds are fixed throughout: seed = 123 for R scripts, seed = 0 for Python background sampling.

Reference

Rossini, et al. (2026). Multi-omic dissection of RNA control reveals convergent cis-regulatory programs during glucose starvation in yeast. BioProject 1022817.

Citation

If you use this code, please cite the associated manuscript (currently under review).