Improvements af

[Ale-pinto-alba]

• Create a method to get the allele frequency from the GnomAD's VCF file. Usually, the alleles frequencies are in a tuple of an unique float, but there could be a list, so I tried to manage all the possibilities, even if it does not have an AF (None).

• Change the assertions to exceptions in the prepare_alt_seq function. I think the past AssertionError was too prohibitive and when I was plotting the results, I found impossible to continue with them. So, I thought it handling them as ExceptionError was a better way to proceed.

• Make personalized errors messages for better readability and update corresponding error test.

[ielis]

Hi,

thanks a lot for the code!

I have two comments, please see below. The get_allele_frequency might need to be updated, and pls consider the other comment as well. Apart from these issues, the code looks good to me.

[Ale-pinto-alba]

Yes, I agree with you.

On the one hand, I decided to go through the possible alternatives alleles in case there is more than one, keeping the index stored and then splicing the corresponding AF tuple with it. I am assuming the AFs are in the same order as the alleles alts.

On the other hand, I created a separate function to check the errors separately, I kept the Reference alleles do not match in the function because I think it is easier this way and I consider it should be checked too because there are some variants that don't fit.

[Ale-pinto-alba]

It's true that can induce bugs, I didn't realize, so I just kept the second part of the method. Making sure that the alternative allele given in the VariantCoordinates is in the VCF file, independently if it is a tuple with only one alternative allele or with several. Therefore, returning the corresponding AF.