functional_annotation.md

Functional Annotation with EnTAP (v.1.3.0)

• Run Parameters:
  • entap_v1.3.0_params
  • e-value: 1e-05
  • query coverage: 50
  • target coverage: 50
• Configuration:
  • entap_v1.3.0_config.ini
• Databases Searched:
  • RefSeq: complete.protein.faa.216.dmnd
  • UniProt: uniprot_sprot.fa.2.0.6.dmnd
• Contaminants Flagged:
  • Bacteria
• Dependency Versions:
  • DIAMOND: 2.1.8
  • EggNOG-mapper: 2.1.7

This is a simplified representation of how EnTAP was run using Fistulifera pelliculosa:

EnTAP --runP --entap-ini entap_v1.3.0_config.ini --run-ini entap_v1.3.0_params --out-dir entap_outfiles --input braker.aa --taxon Fistulifera_pelliculosa

The actual script (entap_v1.3.0.sh) loops through all the species.

Convert BRAKER GTF to GFF with AGAT (v.1.4.0)

In order to append functional information using the Parent/ID relationship, the structural annotation needed to be converted into a GFF. Unfortunately, AGAT mishandled pre-leading introns, resulting in misplaced UTRs/exons. These were removed.

agat_convert_sp_gxf2gxf.pl -g ${gtf} -o ${id}.gff

To parallelize this process, Nextflow was leveraged, see:

• agat.sh
• agat.nf
• nextflow.config

Decorate GFF File with Functional Information

NCBI Datasets and pdb_pfam_mapping.txt are required to decorate the GFF. Instead of using the RefSeq accession (XP*) for dbxref, we pull the GeneID. Likewise, the PFAM accession is needed instead of the name.

python ncbi_gff.py entap_results.tsv Fistulifera_pelliculosa.gff pdb_pfam_mapping.txt Fistulifera_pelliculosa_ncbi.gff

This script will append the following to the 9th column:

• gene_biotype=protein_coding [gene]
• product [mRNA, CDS]
• Dbxref
  • GeneID [all features]
  • UniProtKB/Swiss-Prot [all features]
  • PFAM [mRNA, CDS]
  • GO [mRNA, CDS]
• note=EggNOG description [mRNA, CDS]

The full script (loop.sh) iterates through all the species.