[Inparanoid] InParanoid 4.1

[Erik Sonnhammer]

Yes, I would think so.

/Erik

On 05/08/2015 09:54 AM, Timon Hick wrote:
> Hi,
>
> tanks for your reply. I just checked the version of my blastall and its
> 2.2.26. So I assume, that the differences can be considered as "normal",
> right?
>
> Best Wishes
>
> Timon
>
> On 07.05.2015 15:02, Erik Sonnhammer wrote:
>> Hi,
>>
>> These are not huge differences; this can happen if many homologs are
>> border-line (near 40 bits in score).  I think we have seen such
>> differences between Blast versions - we used blastall 2.2.18. If you are
>> also using that, then I admit it is a bit strange.
>>
>> /Erik
>>
>> On 2015-05-07 10:44, Timon Hick wrote:
>>> Dear Erik,
>>>
>>> thank you for your very fast response.
>>> Some of the differences were indeed very small, like scores differing
>>> 1 percent, but I already thought that to be no major problem.
>>>
>>> I have just performed another run of InParanoid on B. bovis against M.
>>> tuberculosis and obtained yet another result, which is still different
>>> to the one, from your website. In numbers:
>>>
>>> Server-Version:
>>> 207 groups of orthologs
>>> 241 in-paralogs from B. bovis
>>> 275 in-paralogs from M. tuberculosis
>>>
>>> local version:
>>> 204 groups of orthologs
>>> 235 in-paralogs from B. bovis
>>> 268 in-paralogs from M. tuberculosis
>>>
>>> In a previous run, there were less than 200 groups of orthologs.
>>>
>>> What is the reason, that every run yields a different result?
>>>
>>> And are these differences to be considered as big?
>>>
>>> If you need further information to find the problem, let me know.
>>>
>>> Thank you again for your response and your help.
>>>
>>> Timon Hick
>>>
>>> Am 06.05.2015 um 17:24 schrieb Erik Sonnhammer:
>>>> Dear Timon,
>>>>
>>>> there can be small differences between the on-line and regenerated
>>>> InParanoid orthologs, but if there are big ones then probably something
>>>> went wrong.
>>>>
>>>> Small differences may come from different blast versions, differences in
>>>> the InParanoid script that we ran (a speed-optimised version for compute
>>>> clusters), and arithmetic differences between operating systems.
>>>>
>>>> We have archived the blast outputs but it would be a lot of work to dig
>>>> them up.
>>>>
>>>> In terms of nr of orthologs, how different are the results?
>>>>
>>>> Erik Sonnhammer
>>>>
>>>> On 2015-05-06 15:38, Timon Hick wrote:
>>>>> Dear InParanoid-Mailinglist,
>>>>>
>>>>> I am currently working on a project for which I need a bunch of
>>>>> orthology information and I found your database/program to fit my
>>>>> needs.
>>>>> In order to elucidate whether I should use the orthology data available
>>>>> in the download section of your website or whether I should run the
>>>>> program locally, some questions occured, which I wanted to post here,
>>>>> hoping to get professional advice.
>>>>>
>>>>> I subscribed for the standalone version 4.1 of InParanoid and ran it
>>>>> (without changes in the script) on two sequence files, which I also got
>>>>> from your website (v. 8.0, processed). I also loaded the respective
>>>>> archive from your website and compared the sql-tables from both
>>>>> outputs.
>>>>>
>>>>> As far as I am informed, the othology data available on your website
>>>>> were obtained using the same version as me, so I thought the results
>>>>> should be the same, but they were not.
>>>>>
>>>>> So my question is, if you know, where these differences could come
>>>>> from.
>>>>>
>>>>> Another question is, whether the intermediate files (BLAST-outputs) of
>>>>> the runs from which the online available results came from, are
>>>>> available too.
>>>>>
>>>>> I really hope to get help from you and thank you in advance for every
>>>>> answer.
>>>>>
>>>>> Best Wishes
>>>>>
>>>>> Timon
>>>>> _______________________________________________
>>>>> InParanoid mailing list
>>>>> InParanoid at lists.su.se
>>>>> https://lists.su.se/mailman/listinfo/inparanoid-at-sbc.su.se