COVID19 Genome Analysis

feel free to contact evancresswell@gmail.com or <vipulp@niddk.nih.gov > regarding questions on implementation and execution of repo

Package Requirements

• Singularity - to run the proper enviornment for code through the singularity container LADER.simg (https://hub.docker.com/layers/123851957/evancresswell/erdca/LADER/images/sha256-bf25a591c1682fa1c88eb7cd1b7b9dcbb13848562a0450b5aebfd0b2994e6241?context=explore)
• MAFFT - to align genome sequences

Table of Contents

• Genome sequence Alignment
  • Align full genome
  • extract clades from full alignment
• Infer interactions using Expectation Reflection
• Post-processing and data visualizaion
  • plotting of interaction maps
  • clade incidence plotting
  • amino acid and nucleotide analysis
• Results
  • Result text files used for paper figures

Genome sequence alignment

Back to Top

Required Files

• 09062020-Cov19gisaid--SeqIdDesc200-aligned-FastaCikti.fa -- previously aligned fasta file (for testing)
  • downloaded from http://www.dilekbalik.com/SARS-CoV-2_ODOTool/
• wuhan_ref_fasta -- covid reference genome: NC_045512.2
• 09062020-Cov19gisaid--SeqIdDesc200-aligned-FastaCikti.fa -- Full un-aligned fasta file
  • downloaded Oct 20 20202 from https://www.epicov.org/epi3/entities/tmp/tmp_sd_2020_10_20_02_12_qigibl_4j510bafde3/sequences_2020-10-19_07-21.fasta.gz

Alignment Process

Back to Top

we want to generate cov_gen_aligned.fasta from our un-aligned fasta file using the Wuhan reference genome

FASTA Data:

All FASTA data both created and required for these simulations can be found in the fasta_files/ directory, organized as follows.

• .fasta files containing full genome and clades alignments

Test run

foo@bar:~$ mafft --auto --keeplength --addfragments 09062020-Cov19gisaid--SeqIdDesc200-aligned-FastaCikti.fa wuhan_ref.fasta > 

Full Genome Alignment

• Break up un-aligned fasta files
  • Assumes you have the following files in /path/to/er_covid19/
    • subject genome file: wuhan_ref.fasta
    • aligned genome file: cov_gen_aligned.fasta
    • directory for output: /path/to/er_covid19/cov_fasta_files/

foo@bar:~$ singularity exec -B /path/to/er_covid19/biowulf,/path/to/er_covid19/covid_proteins /path/to/er_covid19/LADER.simg python break_up_fasta.py /path/to/er_covid19/ 

- This creates 
	- broken up fasta files for alignment in *path/to/er_covid19/cov_fasta_files/* 
	- script to align:  **cov_align.swarm** (for cluster computation)

• Run alignment pieces which aligns with mafft (this can be done on a cluster or sequentially).

foo@bar:~$ ./submit_align_swarm.script 

- Individual imulation requirements
	- batches of 15 (lines in **cov_align.swarm**)
	- 10 GB
	- 2 hours

• Finish by concatenating resulting mini-alignments in cov_fasta_files/ (directory created to house mini-alignments) to create full alignment: covid_genome_full_aligned.fasta

foo@bar:~$ singularity exec -B /path/to/er_covid19/biowulf,/path/to/er_covid19/covid_proteins /path/to/er_covid19/LADER.simg python concat_fasta.py /path/to/er_covid19/ 

Get Clade Alignments

• once you have the full genome aligned file you can get clades using get_clades.py
• get_clades.py has subject file hard coded:

foo@bar:~$  singularity exec -B /path/to/er_covid19/biowulf/,/path/to/er_covid19/covid_proteins /path/to/er_covid19/LADER.simg python get_clades.py /path/to/er_covid19/

Infer interactions using Expectation Reflection

Back to Top - Assumes you have the following files in /path/to/er_covid19/ - subject genome file: wuhan_ref.fasta - aligned genome file: covid_genome_full_aligned.fasta (created in [Full Genome Alignment][#Full-Genome-Alignment]) - directory for output: /path/to/er_covid19/cov_fasta_files/

• once you have an aligned file you can get DI using ER using run_covGENOME_ER.py
• file generates .pickle files with DI
• the different clade and full sequence run are already defined in run_cov_GENOME_ER file.
  • for all clades see submit_DI_clade_swarm.script:

foo@bar:~$ ./submit_DI_clade_swarm.script

- for full genome see **run_covGENOME_ER.py**
	- hardcoded existing full aligned file

foo@bar:~$  singularity exec -B /path/to/er_covid19/biowulf/,/path/to/er_covid19/covid_proteins /path/to/er_covid19/LADER.simg python run_covGENOME_ER.py 

Post-Processing and Data Visualization

Back to Top

Scripts and Drivers

• incidence_plotting.py - file for plotting incidence by region and clade
• gen_incidence_data.py - driver for generating paper incidence data
• make_codon_swarm.py - script to create swarm (cluster) file to map sequences to to resulting amino acids using codon_mapping.py
• make_pair_aa_swarm.py - script to create swarm (cluster) file to map network interactions (from ER simulation) to resulting amino acids using codon_mapping.py
• codon_mapping.py - file which maps a given sequence/region of nucleotides to the resulting amino acid
• print_pair_counts.py - driver for creating lists of amino acid pair counts from ER-inferred position pairs (detailed description in file)
• covid_aa_bp_variance.py - a file which computes statistics of amino acid and basepair expression from given genome positions
• plot_covid_genome_CT.py - file for plotting co-evolving postions of genome by region and clade
• expectation_reflection.py - file to run inference of network interactions using Expectation Reflection

Results

Back to Top All the data generated by the above scripts are in the Results/ directory and are organized as follows.

Result Data:

• .pickle files containing DI of full genome and clades both processed and unprocessed
• .npy files containing amino acid and basepair data for infered genome positions
• .txt files containing all DI pairs from clade and full genome simulations (for inspection without python run)