Bug in checking weights sum to 1

[BeyondTheProof]

Hi Avanti, I've reopened this issue to fix a small bug. You had a < instead of > in your fix. I left a one-line review on the PR.

Originally posted by @BeyondTheProof in #79 (comment)

[AvantiShri]

Thanks, I have a PR for that here #82 and will merge into master once the tests pass. Also version 10 had introduced a new default of variable-length seqlet identification, but I've decided to roll back that default since it doesn't seem as good as the previous fixed-length seqlet identification default. I'll merge it into master when it passes tests.