Potato (Solanum tuberosum) is an autotetraploid, clonally propagated crop and the fourth most important food staple worldwide; late blight, caused by the hemibiotrophic oomycete Phytophthora infestans, remains its most damaging disease. Recent work shows that infection reshapes the potato DNA methylome, but whether these methylation changes causally control immune-gene expression — and whether an infection-induced methylation state is inherited through the vegetative (tuber) cycle rather than reset — is unknown. This project generates and analyses genome-wide DNA methylation and transcriptome data to test these questions in a haplotype-resolved tetraploid background.
The computational work has two phases. Phase 1 (public-data reanalysis) integrates existing whole-genome bisulfite sequencing (WGBS) and RNA-seq from published late-blight and baseline potato studies, plus causal tomato demethylase datasets, against the DM v6.1 reference and the haplotype-resolved cv. Désirée assembly (~0.88 Gb per haplotype set). We call context-specific (CpG/CHG/CHH) differentially methylated regions (DMRs) and differentially expressed genes (DEGs), and test their spatial overlap with the ~2,000 NB-ARC (NLR) immune-receptor loci and flanking transposable elements. Preliminary analysis on baseline data already shows that ~50% of potato NLR genes sit in methylated, TE-flanked ("demethylation-competent") promoters, motivating the causal test.
Phase 2 (own-generated data) analyses WGBS and RNA-seq libraries from a controlled experiment: a CRISPR knockout of the potato ROS1/DME-family DNA demethylase versus wild type, with and without P. infestans infection, tracked across a tuber propagation cycle (mother plant → daughter tuber → regenerated plant). This yields the causal and heritability read-outs at the core of the project.
Both phases are dominated by CPU-intensive short-read alignment and methylation calling. Bisulfite alignment against a large, repeat-rich tetraploid genome is the principal cost; DMR and differential-expression analyses are lighter, memory-bound R workloads that we parallelise per chromosome. All software is standard, open-source, and multithreaded (Trim Galore, Bismark/BatMeth2, HISAT2/STAR, samtools, methylKit/DSS, DESeq2), running as independent per-sample array jobs that scale near-linearly across cores. Storage requirements are driven by raw sequencing reads and alignment intermediates and grow over the project as own-generated libraries are added.
The outcome is the first integrated, allele-aware map of infection-associated DNA methylation dynamics at potato immune loci, and a direct test of whether a demethylation-gated immune state is mitotically heritable through the tuber cycle — a question that clonally propagated crops can address but seed models cannot. The results provide fundamental insight into epigenetic regulation of plant immunity and a basis for durable, input-reducing crop protection strategies.