search

williamslab / ibis

Algorithm for rapid, phase-free detection of long identical by descent segments
GNU General Public License v3.0
17 stars 7 forks source link

Problemi IBS installation #7

Closed Luciagenetic closed 1 year ago

Luciagenetic commented 2 years ago

hi i can't get ibis running maybe i have a problem with the installation. I am to insert the map in cm in the bim file but the following steps do not work

DanielSeidman commented 2 years ago

Hello! Can you describe what happens with the bim file in more detail? Also can you describe the format of the map file you are using? If the genetic positions in the bim file are still 0 after running the insertion script, it may mean the map you are using is in the wrong format, which is the most common problem I see people have with this step. There are multiple different formats for storing genetic map data and the insertion script doesn't read all of them correctly.

Luciagenetic commented 2 years ago

hello thanks for the reply. ibis now works. the problem was in the installation after the make I saw that in the genetio folder in the nuclearfamily.cc file was missing bit I made a change and everything works. the files were formatted correctly with plink and then I entered the map in cm. is there a way to treat inbreeding? thank you

Da: DanielSeidman @.> Inviato: mercoledì 6 luglio 2022 17:47 A: williamslab/ibis @.> Cc: Luciagenetic @.>; Author @.> Oggetto: Re: [williamslab/ibis] Problemi IBS installation (Issue #7)

Hello! Can you describe what happens with the bim file in more detail? Also can you describe the format of the map file you are using? If the genetic positions in the bim file are still 0 after running the insertion script, it may mean the map you are using is in the wrong format, which is the most common problem I see people have with this step. There are multiple different formats for storing genetic map data and the insertion script doesn't read all of them correctly.

— Reply to this email directly, view it on GitHubhttps://github.com/williamslab/ibis/issues/7#issuecomment-1176385126, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AZ5TASGPJXSPCR7OWJ2QACLVSWTCRANCNFSM52WA52XA. You are receiving this because you authored the thread.Message ID: @.***>

DanielSeidman commented 2 years ago

Inbreeding is hard with regards to IBD calling. You end up with longer segments than expected. Depending on your goals, and how much inbreeding is expected, you may want to set a harsher length cutoff for segments of interest, and if you are using the .coef file for degrees of relatedness, you may want to ignore the degrees it gives and use a different set of thresholds based on how much inbreeding is expected. I don't have a perfect answer for you since it's a difficult problem.

Luciagenetic commented 2 years ago

pedigree in graph. thanks

Ottenere Outlook per Androidhttps://aka.ms/AAb9ysg

From: lucia saccuzzo @.> Sent: Wednesday, July 6, 2022 7:15:28 PM To: williamslab/ibis @.>; williamslab/ibis @.> Cc: Author @.> Subject: Re: [williamslab/ibis] Problemi IBS installation (Issue #7)

I have changed the default parameters for both IBD1 and IBD2 as ouput I have done both coef in coef seg and the file for the ROH in this I will have to decrease the markerconsider threshold. the degree of kinship is likely. is there any software that takes these results to build a pedigree? thanks to the availability

Ottenere Outlook per Androidhttps://aka.ms/AAb9ysg

From: DanielSeidman @.> Sent: Wednesday, July 6, 2022 7:08:05 PM To: williamslab/ibis @.> Cc: Luciagenetic @.>; Author @.> Subject: Re: [williamslab/ibis] Problemi IBS installation (Issue #7)

Inbreeding is hard with regards to IBD calling. You end up with longer segments than expected. Depending on your goals, and how much inbreeding is expected, you may want to set a harsher length cutoff for segments of interest, and if you are using the .coef file for degrees of relatedness, you may want to ignore the degrees it gives and use a different set of thresholds based on how much inbreeding is expected. I don't have a perfect answer for you since it's a difficult problem.

— Reply to this email directly, view it on GitHubhttps://github.com/williamslab/ibis/issues/7#issuecomment-1176471858, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AZ5TASBQHDEVG2IGB5L3PQ3VSW4PLANCNFSM52WA52XA. You are receiving this because you authored the thread.Message ID: @.***>

Luciagenetic commented 2 years ago

I have changed the default parameters for both IBD1 and IBD2 as ouput I have done both coef in coef seg and the file for the ROH in this I will have to decrease the markerconsider threshold. the degree of kinship is likely. is there any software that takes these results to build a pedigree? thanks to the availability

Ottenere Outlook per Androidhttps://aka.ms/AAb9ysg

From: DanielSeidman @.> Sent: Wednesday, July 6, 2022 7:08:05 PM To: williamslab/ibis @.> Cc: Luciagenetic @.>; Author @.> Subject: Re: [williamslab/ibis] Problemi IBS installation (Issue #7)

Inbreeding is hard with regards to IBD calling. You end up with longer segments than expected. Depending on your goals, and how much inbreeding is expected, you may want to set a harsher length cutoff for segments of interest, and if you are using the .coef file for degrees of relatedness, you may want to ignore the degrees it gives and use a different set of thresholds based on how much inbreeding is expected. I don't have a perfect answer for you since it's a difficult problem.

— Reply to this email directly, view it on GitHubhttps://github.com/williamslab/ibis/issues/7#issuecomment-1176471858, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AZ5TASBQHDEVG2IGB5L3PQ3VSW4PLANCNFSM52WA52XA. You are receiving this because you authored the thread.Message ID: @.***>