Codes for Correlational networking guides the discovery of cryptic natural product biosynthetic enzymes

This repository is the collection of codes for our paper: Correlational networking guides the discovery of cryptic natural product biosynthetic enzymes

Abstract

Bacterial natural product biosynthetic genes, canonically clustered, have been increasingly found to rely on cryptic enzymes encoded elsewhere in the genome for completion of biosynthesis. The study and application of lanthipeptides are frequently hindered by cryptic proteases required for final maturation. Here, we establish a global correlation network bridging the gap between lanthipeptide precursors and cryptic proteases. Applying our analysis to 161,954 bacterial genomes, we establish 5,209 correlations between precursors and cryptic proteases, with 91 prioritized. We use network predictions and co-expression analysis to reveal a previously missing protease for the maturation of class I lanthipeptide paenilan. We further discover widely distributed bacterial M16B metallopeptidases of previously unclear biological function as a new family of lanthipeptide proteases. We show the involvement of a pair of bifunctional M16B proteases in the production of novel class III lanthipeptides with high substrate specificity. Together, these results demonstrate the strength of our correlational networking approach to the discovery of cryptic lanthipeptide proteases and potentially other missing enzymes for natural products biosynthesis.

Dependencies

In order to run the scripts, you need to install Python 3.7+, R 3.6+, pandas and openpyxl. You could install them using conda:

conda install R pandas openpyxl

Pipeline

This section shows how to reproduce correlation analysis in the paper.

Data

To reproduce the results in paper, you should download supplementary file 1.xlsx and 23777967_protease_cluster.csv.xz from Zenodo to the same directory of the scripts. Then, extract the dataset by xz:

xz -d 23777967_protease_cluster.csv.xz

Create count matrix

Run the script create_count_matrix.py to prepare input matrix (require ~64 GB RAM):

python create_count_matrix.py

This will generate a csv file df_count.csv

Construct correlation network

After getting df_count.csv, run spearman_network_taxonomy.R to calculate correlations between proteases and precursors:

Rscript spearman_network_taxonomy.R

This will generate 5 files:

• correlation_spearman_genus.csv is the raw correlation result
• Fig1_1_genus.csv is the filtered result corresponding to Fig. 1B of the paper
• network_spearman_genus.csv, node_spearman_genus.csv, edge_spearman_genus.csv are network files of filtered result (ρ>0.3, pAdj<1E-5, I>=10) that can be visualized by Cytoscape.