TE Pipeline

Snakemake pipeline that runs RepeatMasker on GENCODE transcripts, extracts TE features per transcript, and compares distributions between protein-coding and lncRNA classes via statistical tests and plots.

Requirements

• Conda / Mamba (environment: workflow/envs/te_analysis.yaml)
• RepeatMasker (installed in the conda env)
• Snakemake ≥ 9

Quick start

# copy and edit the default config
cp config/config_v47.yaml config/config_mine.yaml
# edit paths: gencode_gtf, gencode_fasta, pc/lncrna FASTAs or ID files

snakemake --snakefile Snakefile \
          --configfile config/config_mine.yaml \
          --profile profiles/default \
          --use-conda

# or simply
snakemake

Classification modes

Set classification_mode in the config to one of:

Mode	Required keys
fasta	pc_transcripts_fasta, lncrna_transcripts_fasta
id_file	pc_transcript_ids_file, lncrna_transcript_ids_file

Pipeline steps

Step	Rule	Output
1	parse_gencode_gtf	annotation/transcripts_from_gtf.bed, transcript_biotypes.txt
1.5	extract_transcript_lengths	annotation/transcript_lengths.txt
2	extract_ids_from_fasta	annotation/pc_transcript_ids.txt, lncrna_transcript_ids.txt
4	index_transcripts	annotation/all_transcripts.fa.fai
5	check_fasta_headers	annotation/all_transcripts_headers_checked.fa
5	run_repeatmasker_full	repeatmasker/all_transcripts.out{,.gff}
6	extract_all_features	features/all_transcripts_te_features.csv
7	filter_features_by_class	features/{pc,lncrna}_transcripts_te_features.csv
8	combine_classified_features	combined/classified_te_features.csv
9	statistical_analysis	analysis/univariate_tests.csv, summary_report.txt, …
10	generate_visualizations	plots/te_presence_comparison.png, volcano_plot.png, pca_plot.png

All outputs are namespaced under results/{dataset}/.

Configs

File	Purpose
config/config_47.yaml	GENCODE v47 run
config/config_49.yaml	GENCODE v49 run

Dev info

See te_pipeline_manifest.yaml for the full rule catalogue, script inventory, dataset run status, and cleanup candidates.