Assembling a corpus of phosphoproteomic annotations using ProtMapper to normalize site information from databases and text mining
This repository contains scripts and other files for generating the results of the manuscript. The results build on both the ProtMapper and INDRA systems.
ProtMapper is available at: https://github.com/gyorilab/protmapper INDRA is available at: https://github.com/sorgerlab/indra
The Makefile contains all the commands for generating results based on
input data. The scripts used for processing are inside the protmapper_paper
package.
The protmapper_paper/get_sites module contains scripts to extract
phosphosite annotation data from various sources including:
- Machine reading-based annotations aggregated by INDRA using the Reach, Sparser and RLIMS-P systems.
- Annotations from the SIGNOR, HPRD, Reactome, NCI-PID, Panther and PhopsphoSitePlus databases using source processors implemented in INDRA.
The protmapper_paper/map_sites.py script uses the ProtMapper to
map all "raw" sites collected from the above sources.
The following scripts in the protmapper_paper module perform analysis on the
aggregated and mapped phosphosite annotation data:
analyze_sites.py: Statistics on the number of sites and annotations from different sources.annotation_count.py: Comparing the number of sites and annotations from different sources and the effect of applying ProtMapper mapping in the context of analyzing phosphoproteomic data.annotatioon_stmts.py: Construct INDRA Statements representing phosphosite annotations in the assembled corpus. This is the basis for the export.mapping_stats.py: Statistics on the effect of applying ProtMapper mapping to phosphosite annotations aggregated from different sources.psp_reading_venn: Create Venn diagrams of overlap between sites and annotations from different sources.sample_reading_sites.py: Script to sample sites from machine reading for curation.belief.py: Script to calculate INDRA belief score corresponding to phosphosite annotations in the assembled corpus.