AnnotationDB Integration Required Mods Bugs - NIAGADS/GenomicsDBWebsite GitHub Wiki
EGA Modified .tsx/.ts files
-
MostSevereConsequences.tsx
-- probably no additional mods needed at this time- updated to reflect new attribute names and types
- layout altered
-
types.ts
-- cleanup still needed- updated to add/remove
VariantRecordAttributes
- clean up still needed for this and other record attributes (probably should wait until other record types are fixed)
- updated
GeneRecordAttributes
- updated to add/remove
-
VariantRecordHeading.tsx
-- probably no additional mods needed at this time- removed references to removed attributes
- refactored layout, especially
ADSPQCDisplay
component
-
GeneRecordHeading.tsx
- updated attribute names for genetic evidence for ad flags
- commented out embeded igv browser
Website Bugs?! / Needed Fixes
-
record
visible (as defined inindividuals.txt
) attributes are assumed to be tables and throw an error precluding page load- e.g.,
cadd_phred_score
-- currently set torecord-internal
inindividuals.txt
to allow page load; will need to alter that torecord
to debug
- e.g.,
-
Record Tables
- "No variants meet the default p-value cutoff for genome-wide significance..." when you filter on text phrase that is not matched; not sure if that is the behavior we want; should discuss (fixed in 15f82d4)
- inline bar-charts (e.g., allele frequencies) -- switch the colors, making the lighter one the values -- (fixed in initial branch integration)
-
Motif Consequences
table rendering error- see:
app/record/variant/1:68742147:A:G_rs542665189#motif_consequences
uncaught Error: Objects are not valid as a React child (found: object with keys {icon, type, color, value}). If you meant to render a collection of children, use an array instead.`
- see:
`in div (created by TdComponent)` `in TdComponent (created by ReactTable)` `in div (created by TrComponent)` `in TrComponent (created by ReactTable)` `in div (created by TrGroupComponent)` `in TrGroupComponent (created by ReactTable)` `in div (created by Tbody)
* problem may be related to `motif_score_change`; uncertain
**UPDATE:** it seems to be a problem with the attribute type, which right now is set to `numeric` but, judging by the payload, should be `json_icon` -CK
Transcript Consequences
table on Variant Page- Coding? check is not showing (this is probably an issue in the query)
- results are supposed to be ordered by rank ASC; check qeury
DB Bugs / Need Fixes
-
Tuning Indexes
- to speed up LocusZoom GWASService: Results.VariantGWAS on protocol_app_node_id, 'chr' || split_part(variant_record_primary_key, ':',1)::text, and possibly position
-
NIAGADS.Variant
- sequence alleles for deletions may be wrong b/c length check was based
on alt instead of norm alt. corrected in
VariantAnnotator
PERL lib and in real-time generated alleles from AnnotatedVDB, need to patch db (both sequence and display alleles)- can possible use
display_allele_attributes('GTTT','TT', 'GT', 'T',196994570)
SQL function to facilitate patching
- can possible use
- sequence alleles for deletions may be wrong b/c length check was based
on alt instead of norm alt. corrected in
-
Missing
SRes.ExternalDatabases
- need to add in external datatabase to the GenomicsDB for
gnomAD
andESP
so can join on version for allele frequency table
- need to add in external datatabase to the GenomicsDB for
-
AnnotatedVDB
-
ea
/aa
in allele frequencies are not from 1000genomes but fromNHLBI-Exome Sequencing Project (ESP)
- patch AnnotationDB by iterating over and fixing allele_frequencies field
- correct loading scripts
- will need to update website queries, which currently correct the error in real-time
-
update loading scripts to do lookups of internal Ensembl motif ids (motif regulatory consequences) during load so do not have to require storage of huge motif database (current solution)
-
-
Future review
- NHGRI GWAS SNP x SNP interactions from the NHGRI GWAS catalog (currently skipped)