DependentMutations

This repository contains the series of Jupyter notebooks used to generate the results for Green et al, 2021, https://www.biorxiv.org/content/10.1101/2022.06.24.497504v1

Input and Output

Input data files and resulting output are quite large and are hosted separately at Harvard Dataverse: https://doi.org/10.7910/DVN/KMAACV This also contains input and output from running linear mixed models in GEMMA

Codebase

The notebooks are run in numerical order.

• 00.convert_tree_data_to_event_matrices.ipynb - converts the SNPPar output data into matrices of L x N, where L is the number of branches and N is the number of SNPs. Computes a matrix of mutation events, reversions events, and genetic background, for each lineage individually and then combines.

• 01.event_matrices_to_pairs.ipynb - Computes the first-order (single site) number of mutations. Computes matrix of all possible pairs of mutations. Determines which pairs of mutations are found to ever co-occur.

• 02A.compute_p_values_sequential.ipynb - For pairs of mutations that actually co-occurr on sequential branches, computes the p-value and Benjamini-Hochberg multiple testing correction.

• 02A.compute_p_values_sequential.ipynb - For pairs of mutations that actually co-occurr on simultaneous branches, computes the p-value and Benjamini-Hochberg multiple testing correction.

• 03.single_mutation_annotation.ipynb - Annotates all homoplastic SNPs with information about function, lineage, etc. Calculates enrichment statistics for paper

• 04A.analyze_pvalues_sequential.ipynb - analysis of significant, sequentially ocurring dependent mutation pairs

• 04B.analyze_pvalues_simultaneous.ipynb - analysis of significant, simultaneously ocurring dependent mutation pairs

• 05.analyze_antibiotic_hits.ipynb - analyze significant dependent mutations pairs related to antibiotic resistance

• 06.antibiotic_potentiators.ipynb - determination of which hits "potentiate" antibiotic resistance

• 07.antibiotic_potentiators_table.ipynb - creates tables of hits that potentiate antibiotic resistance for input to GEMMA

Running GEMMA analyses

Citations

Note that some of the input data is from other publications that should be cited accordingly:

• Comas et al, Nat Gen, 2013 10.1038/ng.1038 (M. tuberculosis pan-susceptible ancestor sequence)

• Vargas et al, 2022, https://doi.org/10.1101/2022.06.10.495637 (phylogenies and SNPpar output ancestral sequence reconstructions)

• Freschi et al, 2021, https://doi.org/10.1038/s41467-021-26248-1 (lineage SNP barcode)

• Coll et al, 2014, https://doi.org/10.1038/ncomms5812 (lineage SNP barcode)

• Kapopoulou A, Lew JM, Cole ST, Tuberculosis (Edinb), 2011. 10.1016/j.tube.2010.09.006 (Gene location data)

• Minato et al, 2019, 10.1128/mSystems.00070-19, (Gene essentiality data)

• WHO catalog of resistance variants, https://apps.who.int/iris/handle/10665/341906