FDC-Mech

Pipeline and analysis workspace for the cefiderocol resistance study in Klebsiella pneumoniae.

This repository captures the code and supporting notes used to study the molecular basis of cefiderocol resistance across 33 local clinical isolates and larger public-genome comparison sets. The workflow combines whole-genome sequencing, protein-level alignment, Prokka/BLAST/Miniprot annotation, LC-MS uptake analysis, population analysis profiling, phylogenetics, and publication figure generation.

What This Repository Contains

• Local clinical isolate analysis for 33 K. pneumoniae genomes.
• Public-genome acquisition and filtering scripts for BV-BRC, NCBI Assembly, and Long et al. comparison sets.
• Resistance mechanism annotation scripts for CirA, OmpK35/OmpK36, FEC operon status, and carbapenemase context.
• LC-MS and PAP analysis scripts for cefiderocol uptake and heteroresistance.
• Phylogenetic and figure-generation scripts for the manuscript and publication figures.
• Working notes and pipeline logs that document the current study state.

Main Study Focus

The analysis centers on how cefiderocol resistance arises in K. pneumoniae, with emphasis on:

• CirA disruption, truncation, and frameshifts.
• Carbapenemase background, especially KPC and NDM combinations.
• Porin loss or alteration in OmpK35/OmpK36.
• Efflux and FEC operon context.
• Phenotype concordance across in-house BMD, commercial systems, LC-MS uptake, PAP, and WGS.

Repository Layout

• analysis/ - core analysis scripts, manuscript drafts, reference protein extraction, and resistance-annotation utilities.
• analysis/Scripts/ - ordered pipeline scripts for public genome retrieval, QC, Prokka, BLAST, Miniprot, phylogenetics, and statistics.
• publication_figures/scripts/ - figure-focused scripts for the manuscript and presentation-ready plots.
• PIPELINE_COMMANDS.md - command-by-command record of the analysis workflow.
• STUDY_PROGRESS.md - live study notebook describing dataset status and interpretation.
• Fasta files/, Dataset/, LCMS data/, Papers/, Presentation/, results/ - local working folders for data, notes, and outputs.

Large generated outputs and local working directories are intentionally excluded from version control via .gitignore.

Key Pipeline Stages

The analysis pipeline is organized around these steps:

1. Build a reference protein set from ATCC 13883.
2. Annotate genomes with Miniprot and Prokka.
3. Search CirA with BLAST+ to classify complete versus incomplete protein hits.
4. Parse protein alignments into mutation summaries.
5. Enrich the master table with Kleborate, MIC, and clinical metadata.
6. Generate iTOL annotation tracks and phylogenetic trees.
7. Run LC-MS uptake, PAP, and statistical analyses.
8. Produce publication figures.

Representative scripts:

• analysis/enrich_merged_table.py
• analysis/generate_itol_annotations.py
• analysis/plot_tree.R
• analysis/select_representatives.py
• analysis/Scripts/01a_bvbrc_specialty_genes.py
• analysis/Scripts/03_download_genomes.py
• analysis/Scripts/04_qc_filter.py
• analysis/Scripts/05_run_prokka_batch.sh
• analysis/Scripts/06_run_blastp_batch.sh
• analysis/Scripts/07_run_miniprot_batch.sh
• analysis/Scripts/08_parse_miniprot_public.py
• analysis/Scripts/09_merge_combined_global.py
• analysis/Scripts/10_snippy_batch.sh
• analysis/Scripts/12_snippy_core_iqtree.sh
• analysis/Scripts/13_statistical_analysis.R
• analysis/Scripts/14_publication_figures.py
• analysis/Scripts/15_enrich_clinical_and_itol.py
• analysis/Scripts/16_fec_operon_census.py
• analysis/Scripts/17_lcms_uptake_analysis.py
• analysis/Scripts/18_pap_analysis.py

Publication-specific scripts are mirrored in publication_figures/scripts/.

Reproducible Inputs And Outputs

The repository is designed to work with the local study data already present in the workspace, including:

• 33 local FASTA assemblies in Fasta files/fasta/.
• Clinical and MIC spreadsheets in Dataset/.
• LC-MS cleaned data in LCMS data/data_clean.csv.
• Reference proteins and alignment outputs in analysis/.

Generated deliverables are written to locations such as:

• analysis/results/
• analysis/results/stats/
• analysis/phylogenetics/
• publication_figures/figures/

Notes

• This repo contains analysis code and study notes, not a packaged command-line application.
• Several folders in the workspace are intentionally not tracked because they contain large intermediate data, raw assemblies, or local-only working files.
• The current study log and pipeline command file are the best starting points if you want to reproduce or extend the analysis.

Suggested Starting Points

If you are trying to understand the project quickly, read these files in order:

1. STUDY_PROGRESS.md
2. PIPELINE_COMMANDS.md
3. analysis/Scripts/09_merge_combined_global.py
4. analysis/Scripts/15_enrich_clinical_and_itol.py
5. analysis/Scripts/17_lcms_uptake_analysis.py
6. analysis/Scripts/18_pap_analysis.py