Re: Competing arrow tools in pbgenomicconsensus and unanimity
On October 11, 2018 11:14:35 AM EDT, Andreas Tille <tille@debian.org> wrote:
>Hi Afif,
>
>On Thu, Oct 11, 2018 at 09:27:16AM -0400, Afif Elghraoui wrote:
>> >
>> >#!/bin/sh
>> >variantCaller --algorithm=arrow $*
>> >
>> >while unanimity installs a compiled binary. @Afif (or whoever is
>> >informed about this PB programs): Do you have a sensible suggestion
>> >which arrow we should install?
>> >
>>
>> At one point, I had read that unanimity was eventually going to
>supersede pbgenomicconsensus altogether. Looking at the source tree
>[3], it looks like genomicconsensus in unanimity is still marked
>experimental, so I would stick with pbgenomicconsensus' implementation.
>>
>> As for the name, "arrow" succeeds "quiver", so there's a little theme
>going on. And quiver had some meaning as a Quality Value-aware variant
>caller.
>
>My conclusion from your answer is that it makes sense to ship arrow
>from
>unanimity while removing the little wrapper from pbgenomicconsensus.
>This can be explained in d/NEWS.Debian.
>
>Do you agree with this conclusion?
>
No, that's the opposite of what I said. The unanimity implementation looks like it's still experimental.
Afif
Reply to: