How to fix truncated BAM files from BWA?

Understand the Issue

• Truncated BAM files often result from incomplete file transfers or interruptions during the BWA alignment process.

 

• BAM files are binary, so corruption isn't always visually apparent. They may fail during downstream processing or when converting to SAM format.

Diagnostic Check

• Use `samtools quickcheck` to identify issues with the BAM file without reading all contents. It checks for termination issues.

 

• Check the file header using `samtools view -H yourfile.bam`. Ensure the header loads correctly.

Repair the BAM File

• Use `samtools` to attempt repairs. The command `samtools quickcheck -r yourfile.bam` can sometimes recover partially written BAM files.

 

• Convert BAM to SAM using `samtools view yourfile.bam > yourfile.sam`. This process sometimes circumvents issues in reading the BAM file.

Re-create the BAM File

• If repairs fail, regenerate the BAM file from raw data. Rerun the BWA alignment process from original FASTQ files, if available.

 

• Ensure sufficient storage space and stable system resources during re-creation to prevent future truncations.

Preventive Measures

• Maintain a robust data backup and integrity check routine. Use checksum verification to ensure file integrity before and after file transfers.

 

• Regularly update `samtools` and BWA to the latest versions to benefit from improvements in handling such issues.