SecStrAnnotator:Analysis

SecStrAnnotator Suite provides scripts (Python, R) for batch annotation of the whole family and analysis of the annotation results.

Procedure

Data preparation

The directory scripts/secstrapi_data_preparation/ contains a pipeline for annotating the whole protein family, including:

downloading the list of family members defined by CATH and Pfam,
downloading their structures,
selecting a non-redundant set,
annotation,
multiple sequence alignment for individual SSEs,
formatting into SecStrAPI format,
formatting into TSV format for further analyses.

The whole pipeline can be executed by scripts/SecStrAPI_pipeline.py

Example usage:

 python3 scripts/SecStrAPI_pipeline.py scripts/SecStrAPI_pipeline_settings.json --resume

Before running, modify the settings in SecStrAPI_pipeline_settings.json to set your family of interest, annotation template, data directory etc (see README.txt for more details).

Data analysis

The directory scripts/R_sec_str_anatomy_analysis/ contains a pipeline for statistical analysis of the annotation results on the whole protein family, including:

reading the annotation results from TSV,
generating plots,
performing statistical test to compare eukaryotic and bacterial structures (or any two sets of structures).

Example usage:

Launch rstudio from the said directory
In sec_str_anatomy.R, set DATADIR to the path to your annotation data created in #Data preparation
In sec_str_anatomy_settings.R, modify the family-specific settings (list of helices and strands)
Run sec_str_anatomy.R line by line

Example case study: Cytochromes P450

Data

For the Cytochrome P450 family, structures of 1855 protein domains are available, located in 1012 PDB entries (updated on 7 July 2020). The analysis was performed on a non-redundant subset containing 183 protein domains.

The data are available here (structural files not included because of their size).

Occurrence of SSEs

The occurrence describes in what percentage of the structures a particular SSE is present.

Occurrence of particular SSEs in the whole set.
Occurrence of particular SSEs – comparison of bacterial and eukaryotic structures.

Length of SSEs

The length of an SSE is measured as the number of residues. The following violin plots show the distribution of length for each SSE.

Length distribution of particular SSEs in the whole set.
Length distribution of particular SSEs – comparison of bacterial and eukaryotic structures.

Sequence of SSEs

The amino acid sequences for each SSE can be aligned and used to produce a sequence logo. Where the sequence conservation is sufficient, we can establish a generic numbering scheme: the most conserved residue in helix X serves as its reference residue and is numbered as @X.50. The remaining residues in the helix are numbered accordingly.