Hi all,
I'm working with Illumina GSA (Global Screening Array) microarray data and have a question regarding genotype calling when using different manifest files.
I ran the same IDAT files through the analysis pipeline using two different manifest files. The only differences between these manifests are the chromosome and genomic position annotations — the probe IDs, probe sequences, and other fields remained identical.
My assumption is that the chr and position information in the manifest are only used for annotation, and therefore should not affect genotype calling, which is driven by signal intensity clustering (e.g., via the GenCall algorithm in GenomeStudio or similar).
However, I noticed that some genotype calls changed between the runs. For example:
- A homozygous call changed to heterozygous, or vice versa.
- Some SNPs previously called as '0' (no call) were later assigned genotypes, and vice versa.
This raises a few questions:
- Can differences in the manifest file that are limited to chromosome and position fields really cause such changes in genotype calls?
- Or is it more likely that the changes are due to technical issues, such as a wet-lab error, or mistakes in re-generating result files using different manifest versions?
If anyone has experience with this or can confirm how the manifest impacts genotype calling, I’d really appreciate your input.
Thanks in advance!
