fixed #799 #805
fixed #799 #805
Conversation
|
@pieterlukasse could you help to review? |
|
@jjgao : nice to see there is a class to import genes. Regarding microRNAs, maybe you should also update the class ImportMicroRNAIDs in a similar way? |
|
@pieterlukasse: @sheridancbio is working on better support for miRNA.. we will update ImportMicroRNAIDs if necessary. |
|
@jjgao : I meant the specific change where you read the symbol from column 10 instead of column 2. This same mistake (i.e. column 2) is done in ImportMicroRNAIDs. |
|
@pieterlukasse ImportMicroRNAIDs is using another file, not the NCBI gene file. |
…on.countSamplesWithProteinPosStarts() - now construct geneId set from proteinPosStarts argument, and use for WHERE clause on table 'mutation' - previously temporary table before join might include almost all mutations (depending on supplied geneticProfileId argument) - changed join to an explicit INNER JOIN (from comma operator join)
|
@jjgao : sorry, I realised this later. Indeed another file :) Which file/service are you using by the way? |
|
@pieterlukasse: It's a checked in pre-cursor and mature miRNA ID mapping file. Rob is writing some script to generate that file automatically from mirbase. |
|
@jjgao I would like to suggest we add the following constraint to DB to correctly reflect the fact that Hugo symbols should be unique. alter table
cbioportal.gene
add
UNIQUE KEY `UQ_HUGO_GENE_SYMBOL` (`HUGO_GENE_SYMBOL`) COMMENT 'Constraint to ensure we have unique hugo symbols'; |
|
Yes, the more constraints the better :-) |
|
@jjgao I tested importGenes.pl (as this is reported in as the "usage" string). In this commit you can find the changes I made to get it working: thehyve@f444ca0 One remark about the logging: I think it is confusing to see this message at the end: while in fact no records were inserted (all were already present in DB). Maybe this is another issue that needs to be logged in a new ticket? |
…omutation_fix Fix DaoMutation call countSamplesWithProteinPosStarts to help with mysql slowdown
patient view performance: hide cbioportal column in mutations table & remove mut vs cna plot
|
@jjgao : I loaded the new NCBI file using the fixed gene importer and can confirm that hugo column is now unique. In fact, I successfully applied the UQ constraint (which I mentioned before) to my DB. Could you add this UQ constraint to this PR as well? |
|
@jjgao : I see that my proposed fixes in previous comments were not taken up. Would you like me to send them via a separate PR? |
This pull request is to fix issue #799.