Fail Import When Data Genome Build Does Not Match Study Genome Build

[forus]

Summary

Some data types (e.g., MAF, SV, others) include a reference genome build field such as NCBI_Build. Today, the importer only logs a warning when the genome build in the file does not match the study’s configured genome build. This can lead to silent inconsistencies or downstream errors.

This issue proposes enforcing strict validation: if a record’s genome build does not match the study’s genome build, the importer should throw an error and stop the load instead of continuing.

Problem Description

Many genomics file formats define a reference genome build. For example, in ImportExtendedMutationData, the loader currently checks the genome build like this:

if (!record.getNcbiBuild().equalsIgnoreCase(genomeBuildName)) {
    ProgressMonitor.logWarning("Genome Build Name does not match, expecting " + genomeBuildName);
}

For all genome-build–aware datatypes (e.g., MAF, SV, others), the importer should fail fast.

[devsisodiya931]

Some mutation files (like MAF) contain a column called NCBI_Build which tells the genome build used in the file.

Earlier in cBioPortal:

If genome build in the file did not match the study genome build

The importer only showed a warning

Import still continued

This is risky because:

Wrong genome build data can be imported

It may cause wrong analysis later

So the requirement was:

If genome build does not match → stop the import immediately

File I Worked On

I searched where mutation (MAF) files are imported.

Main file found:
org.mskcc.cbio.portal.scripts.ImportExtendedMutationData

This file is responsible for:

Reading MAF files

Parsing mutation records

Importing them into database

Old Code (Before Fix)

Earlier, genome build was checked like this:

if (!record.getNcbiBuild().equalsIgnoreCase(genomeBuildName)) {
ProgressMonitor.logWarning(
"Genome Build Name does not match, expecting " + genomeBuildName
);
}

Problem:

Only a warning was printed

Import did not stop

Wrong data could still be loaded

What I Changed (Main Fix)

I changed this logic to fail fast.

New Code Added
if (record.getNcbiBuild() != null &&
!record.getNcbiBuild().equalsIgnoreCase(genomeBuildName)) {

throw new IllegalArgumentException(
    "Genome build mismatch in file " + mutationFile.getName()
    + ". Expected: " + genomeBuildName
    + ", Found: " + record.getNcbiBuild()
);

}

Result:

If genome build does not match → import stops

No wrong data goes into system

Error message is clear and helpful

Issue I Faced

While testing, I noticed:

Importer was trying to connect to MySQL database first

If DB was not running, a DaoException was thrown

Because of this, genome build validation was not reached

How Meaning I Solved It

I placed genome build validation early in the import loop

Validation now happens before database operations

This ensures:

Correct exception is thrown

Import truly fails fast

Test I Used
Test file:
ImportExtendedMutationDataGenomeBuildTest

What test checks:

MAF file with wrong genome build

Import should throw:

IllegalArgumentException

Test command:
mvn -Dtest=ImportExtendedMutationDataGenomeBuildTest test

Final Outcome (Simple)

Genome build mismatch is no longer ignored

Import stops immediately

Data integrity is protected

Behavior matches problem requirement