Annotation of Loci

This is side repository to use annotation files from Michele F to assign a gene name to each lead variants or loci within the pQTL meta-analysis pipeline project.

Gene Name Annotation Process

This process was designed to determine the gene names for 7,870 variants mapped to different genetic regions associated with various protein sequence IDs.

Step 1: Identifying Gene Names from Annotated Files

• A filename was generated for each variant using seqid, locus, and SNPID columns.
• These filenames were compared against 12,095 annotated files provided by Michele F.
• The grep_gene_lb() algorithm was applied to extract the correct gene names where a match was found.
• Using this approach, 7,815 out of 7,870 variants were successfully annotated.

Step 2: Handling Missing Annotations

• For the 55 loci where no annotation file was found:
  • Each missing SNPID was manually checked in Locus Breaker.
  • If an annotation file was found for the same SNPID but associated with a different locus or seqid, it was used to fill in missing gene names.
  • This step recovered 16 additional SNPIDs, reducing the missing count to 39.
• The remaining 39 SNPIDs still lacked annotation and required additional data from Michele F.

Step 3: Resolving Multiple Gene Names for a Single SNPID

• For SNPIDs associated with multiple gene names:
  • GeneCards.org was used to verify gene aliases.
  • When multiple names were listed, priority was given to the older HGNC-approved gene name.
• This ensured consistency in gene annotation across all variants.

This structured approach maximized gene annotation accuracy while addressing missing data through manual verification and external resources.

Identification of Epitope Effect

Steps to Implement:

• Read the LB file which has path to each annotated TXT file.
• Loop through each row, read the corresponding annotated TXT file, and filter relevant rows.
• Check if the Consequence column contains any epitope effect's consequences.
• If there’s a match, store TRUE and the corresponding SNPID; otherwise, store FALSE and NA.

Finally, we reported 5 new variables along with LB results:

• missing_ld: Yes/No, indicating if the annotation file is missing for the lead SNP at this corresponding locus.
• epitope_effect_all: Yes/No/NA, if any variant in annotated file has an epitope effect, regardless of gene match.
• genes_with_epitope_effects: List of all unique gene symbols for which an epitope effect was observed.
• epitope_effect: Yes/No/NA, indicating epitope effect with seqid gene-matching only for rows where cis_or_trans == "cis".
• epitope_causing_variant: List of causal variants in a single row (concatenated with ;) causing epitope effect.