norm

[ying0624]

I have a question regarding the normalized XP-EHH output generated by selscan. The normalized result file contains columns in the following order:
chr, start, end, n_snps, frac_top, frac_bottom, perc_top, perc_bottom, top_score, bottom_score.
Could you please elaborate on the definition of each column, especially columns 5 to 10?
In my output, I observed many entries where perc_top = 1 and perc_bottom = 5, and I am confused about the biological and statistical meaning of such values. Clarification on this point is critical to my subsequent data analysis.
Thank you very much for your time and assistance.

[szpiech]

Hello, These should be described in the manual, please let me know if there is any information missing or if you need some clarification after looking. Reasons why per_top and perc_bottom might have 1/5 in each are likely due to windows with low density of markers creating an extremely compressed distribution of scores, such that the threshold for assigning 1% or 5% is essentially meaningless. We will look into modifying this behavior to avoid confusion, but in the mean time you may need to use larger window sizes.

…

-Zachary

On Thu, Jun 4, 2026 at 9:55 PM ying0624 ***@***.***> wrote: *ying0624* created an issue (szpiech/selscan#165) <#165> I have a question regarding the normalized XP-EHH output generated by selscan. The normalized result file contains columns in the following order: chr, start, end, n_snps, frac_top, frac_bottom, perc_top, perc_bottom, top_score, bottom_score. Could you please elaborate on the definition of each column, especially columns 5 to 10? In my output, I observed many entries where perc_top = 1 and perc_bottom = 5, and I am confused about the biological and statistical meaning of such values. Clarification on this point is critical to my subsequent data analysis. Thank you very much for your time and assistance. — Reply to this email directly, view it on GitHub <#165?email_source=notifications&email_token=ABAKRQSIP5HKCYE4GJYMSL346IR7LA5CNFSL4Z3JMQ5C6L3HNF2C22DVMIXUS43TOVSS6NBVHEZTKMBSGA4TRJTSMVQXG33OVJZXKYTTMNZGSYTFMSSWK5TFNZ2KYZTPN52GK4S7MNWGSY3L>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABAKRQSPGTBSUZMBZODB7HL46IR7LAVCNFSM6AAAAACZ3FXXR2VHI2DSMVQWIX3LMV43ASLTON2WKOZUGU4TGNJQGIYDSOA> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/ABAKRQTW4WLESLLS5X26JTD46IR7LA5CNFSL4Z3JMQ5C6L3HNF2C22DVMIXUS43TOVSS6NBVHEZTKMBSGA4TRJTSMVQXG33OVJZXKYTTMNZGSYTFMSSWK5TFNZ2KUZTPN52GK4S7NFXXG> and Android <https://github.com/notifications/mobile/android/ABAKRQTFNMX6NHFKNSBMZIT46IR7LA5CNFSL4Z3JMQ5C6L3HNF2C22DVMIXUS43TOVSS6NBVHEZTKMBSGA4TRJTSMVQXG33OVJZXKYTTMNZGSYTFMSSWK5TFNZ2K4ZTPN52GK4S7MFXGI4TPNFSA>. Download it today! You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>