Chapter_3_Coexpression

Data preparation

Count table - Go to Sulheim S et al. (2020). Download Data S4, extract average values for all triplicate M145 datapoints for use as input to transcriptominer_thesis.py - note, input file is provided here as average_data.csv. Place in same directory as this script. Genome reference - Download NC_003888 from NCBI in genbank format, place in same directory as this script. BGCs - download genbank files of BGCs of interest and place into bgc_files directory.

Run program

Enter reference genome filename for map_file_name variable (line 31) and count_file_name (line 37). Run transcriptominer_thesis.py (extension settings and other parameters between lines 16 and 50 can be changed according to preferences).

Results

For each BGC in the bgc_files directory, 4 interactive html files will be written to the same directory as this script. Filename formats shown below, bold substrings are dynamic and change depending on the query file/run parameters:

• BGCfilename_unfiltered_cor: Correlation matrix of neighborhood spanning from BGC start - upstream extension to BGC end - downstream extension (default extenions are 20 genes each way).
• BGCfilename_mibig_cor: Correlation matrix of neighborhood spanning from BGC start to BGC end
• BGCfilename_line: Line plot of counts for genes in BGC ignoring extensions. Either target (i) all genes, (ii) all genes of a certain functional annotation - see kinds in line 46, or (iii) user-specified genes - see user_line_cds in line 44.
• BGCfilename_cor_mask: Correlation matrix of neighborhood spanning from BGC start - upstream extension to BGC end - downstream extension (default extenions are 20 genes each way). Correlations below mask threshold are set to 0. Mask threshold is average pairwise correlation of bgc - noise (mask is given in filename).

Note - BGCfilename is the filename of a single bgc file from the bgc_files directory that is undergoing analysis.

Sulheim S, Kumelj T, van Dissel D, Salehzadeh-Yazdi A, Du C, van Wezel GP, Nieselt K, Almaas E, Wentzel A, Kerkhoven EJ. Enzyme-Constrained Models and Omics Analysis of Streptomyces coelicolor Reveal Metabolic Changes that Enhance Heterologous Production. iScience. 2020 Sep 3;23(9):101525. doi: 10.1016/j.isci.2020.101525. PMID: 32942174; PMCID: PMC7501462.