I have been working with some WES files and getting surprising results, in which very rare variants are being found much more often than I would expect. The original files have been aligned but there are single tracks which don't look, from viewing on IGV, like they have been placed in the correct place. I have converted them to fastq files and then re-aligned with the same results. I have included a screenshot from IGV with the same locus from the original alignment. Can anyone give me some direction as to whether this is normal? After I carry out Variant Calling (GATK Best Practices), I am being left with some quite unlikely de novo variants as a result.

Many Thanks for any responses, I will provide further information if required.

Here is a link the the image: