Create a Gene Shet reference data update cmd. #3576
Conversation
| class ShetReferenceDataHandler(ReferenceDataHandler): | ||
|
|
||
| model_cls = GeneShet | ||
| url = 'https://storage.googleapis.com/seqr-reference-data/Shet/Shet_Zeng_2023.tsv' |
There was a problem hiding this comment.
this belongs in the gene_constraint subfolder, as specified in the ticket
| class ShetReferenceDataHandler(ReferenceDataHandler): | ||
|
|
||
| model_cls = GeneShet | ||
| url = 'https://storage.googleapis.com/seqr-reference-data/Shet/Shet_Zeng_2023.tsv' |
There was a problem hiding this comment.
add a comment in this file explaining how that file was generated/ where it came from
| yield { | ||
| 'gene_id': record['ensg'], | ||
| 'shet': float(record['post_mean_shet']), | ||
| 'shet_constrained': bool(int(record['shet_constrained'])), |
There was a problem hiding this comment.
I'm not sure we need this field - we should load the Shet score for any gene we have data for, and we will define a cutoff in seqr for whether or not to flag a gene based on the score cutoff, not a database field. What in the ticket do you anticipate will require this column?
There was a problem hiding this comment.
Hmm, the current cutoff looks like 0.1. The requirement ticket doesn't tell the cutoff explicitly. The new 'LoF constr' tag will be displayed when anyone of LoF, HI, or Shet (this column is true).
There was a problem hiding this comment.
the cutoff is something that we use for display purposes. It should not live in the database in any way, so we should not be loading this
| def parse_record(record): | ||
| yield { | ||
| 'gene_id': record['ensg'], | ||
| 'shet': float(record['post_mean (Shet)']), |
There was a problem hiding this comment.
isn't post_mean the name of the actual statistic, and Shet is the whole method? So shouldn't the score column be called post_mean
There was a problem hiding this comment.
It is the column name of the spreadsheet. I don't know what name is more suitable. Maybe we should ask Lynn.
There was a problem hiding this comment.
the column name is post_mean (Shet). The name of this method is Shet. Therefore, this column is representing the post_mean score for the Shet method. Since the name of the table is Shet, we do not need to capture Shet in the name of the score in the table. Therefore, the name of the score in the table should be post_mean
hanars
requested review from
bpblanken and
hanars
and removed request for
hanars
…t-reference-data-cmd
This is one of the PRs for #3539.