Request for information on VariantRecord instantiation

[blex-max]

Hello,

Could I get some information on why trying to instantiate a VariantRecord like v = VariantRecord() results in TypeError: this class cannot be instantiated from Python? Is this a design choice or something inherently unavoidable?

I'm asking because I'm working on a suite of tools to generate fake data for testing bioinformatics functions/packages/tools - scaling up something I've been doing for individual projects - and it would be really useful to be able to generate VariantRecord objects for this purpose. At the moment, if I am to test a function in this way, I have write the fuction such that any call unwraps the data I require from the VariantRecord object rather than using the object directly. This isn't hugely convenient for a variety of reasons. I can do hacky workarounds via creating temp files and so on, but again, not ideal. Perhaps I've missed something obvious though.

As always, thanks for your work on pysam

[awgymer]

I have found that the best way to get a VariantRecord object is by first creating a VariantHeader and then using header.new_record(...).

The source code indicates that this method is liable to changes but I have found it stable and reliable so far. (Albeit you have to dig a bit to work out how to do some things occasionally)

[blex-max]

@awgymer thanks, this is a good start, I appreciate the help. However, it's not immediately obvious to me how having generated a new record I then access that record as a VariantRecord object. Could you point me in the right direction? Figured it out, you assign the result of .new_record() - I assumed it was an in place operation. I'm still keen to get more insight on this from the devs!