mcanouil / eggla

Early Growth Genetics Longitudinal Analysis.
https://m.canouil.dev/eggla/
Other
2 stars 1 forks source link

Ensure GWAS output is consistent with analysis plan #48

Closed mcanouil closed 2 years ago

mcanouil commented 2 years ago
Column Header Description Format Examples
SNPID SNP/INDEL ID label from the imputation server, or from directly genotyped data Text 4:10229 / rs693
STRAND Orientation of the site to the human genome strand used + (Should be aligned to the forward strand) +
BUILD Build of the genome on which the SNP is orientated Numeric 37
CHR Chromosome on which the SNP resides Numeric 1
POS Position of SNP on Base pairs on human genome build used 34000345
EFFECT_ALLELE Allele for which the effect (BETA) is reported SNP: Capital letter (A,C,G,T) / INDEL: Allele output by imputation server A
NON_EFFECT_ALLELE Other allele at this site SNP: Capital letter (A,C,G,T) / INDEL: Allele output by imputation server G
N Total number of samples analysed at this site Numeric, integer 1234
N0 Number of homozygous samples with zero copies of the EFFECT_ALLELE Numeric, integer or float with 4 digits to the right of the decimal (imputed) 623 / 745.2345
N1 Number of heterozygous samples with one copy of the EFFECT_ALLELE Numeric, integer or float with 4 digits to the right of the decimal (imputed) 623 / 745.2345
N2 Number of homozygous samples with two copies of the EFFECT_ALLELE Numeric, integer or float with 4 digits to the right of the decimal (imputed) 623 / 745.2345
EAF Allele frequency of the EFFECT_ALLELE Frequency with 4 digits to the right of the decimal 0.3546
HWE_P Exact HWE p-value for samples analysed, only if genotyped data is used in analysis Scientific E notation with 4 digits to the right of the decimal (set to missing for imputed SNPs) 1.122E-02 / .
CALL_RATE Genotyping call rate for the SNP (N_successful/N_attempted) Frequency 4 digits to the right of the decimal. Set to 1.000 if imputed 0.9936
BETA Estimate of the effect size Numeric float with 4 digits to the right of the decimal 0.2036
SE Estimated standard error on the estimate of the effect size, uncorrected for genomic control Numeric float with 4 digits to the right of the decimal 0.5611
PVAL P value of the variant association Scientific E notation with 4 digits to the right of the decimal 3.244E-10
IMPUTED Is the SNP/INDEL imputed? 0 = Genotyped / 1 = Imputed 1
INFO_TYPE Type of information provided in the INFO column Text, (set to missing if genotyped) impute_info
INFO Measure of information content for the imputed SNP result Numeric float with 4 digits to the right of the decimal (set to missing if genotyped) 0.4832 / .
mcanouil commented 2 years ago

@burrowsk @annihei For now and by default (use_info = FALSE) "INFO" is no longer extracted from the VCF files in run_eggla_gwas() (9c31a75e0bf8a639d8d6e7b192f8477e453a6fdd). This will speed up computation, but won't produce the above output, at least not all columns.

burrowsk commented 2 years ago

Thanks Mikael, I will do a test tun in ALSPAC. I’m now actually utilising parallel for the GWAS so the total run time is now only ~13 Hrs and not a week haha

BW, Kim


Dr. Kimberley Burrows Senior Research Associate in Epidemiology MRC Integrative Epidemiology Unit Population Health Sciences Bristol Medical School University of Bristol Room OS1 Oakfield House, Oakfield Grove, Bristol, BS8 2BN

Email: @.***

http://www.bristol.ac.uk/integrative-epidemiology/ Profile-Kimberley-Burrowshttps://research-information.bris.ac.uk/en/persons/kimberley-burrows(2513bf22-07b2-44e9-ad72-133c3be48c17).html

A top 3 UK university with leading employers (High Fliers 2017) A top 5 UK university for research (REF 2014) A world top 50 university (QS Ranking 2016)

My pronouns are she/her

From: Mickaël Canouil @.> Sent: 15 September 2022 11:57 To: mcanouil/eggla @.> Cc: Kimberley Burrows @.>; Mention @.> Subject: Re: [mcanouil/eggla] Ensure GWAS output is consistent with analysis plan (Issue #48)

@burrowskhttps://github.com/burrowsk @Aannihei For now and by default (use_info = FALSE) "INFO" is no longer extracted in run_eggla_gwas() (9c31a75https://github.com/mcanouil/eggla/commit/9c31a75e0bf8a639d8d6e7b192f8477e453a6fdd). This will speed up computation, but won't produce the above output.

— Reply to this email directly, view it on GitHubhttps://github.com/mcanouil/eggla/issues/48#issuecomment-1247937178, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ACNZZ2KC6ZM3MGR6XBLYZNLV6L6H3ANCNFSM6AAAAAAQBGJQSQ. You are receiving this because you were mentioned.Message ID: @.**@.>>

mcanouil commented 2 years ago

I'll release eggla v0.13.0 soon, I am waiting for some continuous integration workflows/tests to complete first.

mcanouil commented 2 years ago

Closed with be4fcc767aecaa31e84b9f3b5a5091b8ce71ea93 and #72

burrowsk commented 2 years ago

Cheers, will wait for release.


Dr. Kimberley Burrows Senior Research Associate in Epidemiology MRC Integrative Epidemiology Unit Population Health Sciences Bristol Medical School University of Bristol Room OS1 Oakfield House, Oakfield Grove, Bristol, BS8 2BN

Email: @.***

http://www.bristol.ac.uk/integrative-epidemiology/ Profile-Kimberley-Burrowshttps://research-information.bris.ac.uk/en/persons/kimberley-burrows(2513bf22-07b2-44e9-ad72-133c3be48c17).html

A top 3 UK university with leading employers (High Fliers 2017) A top 5 UK university for research (REF 2014) A world top 50 university (QS Ranking 2016)

My pronouns are she/her

From: Mickaël Canouil @.> Sent: 15 September 2022 12:08 To: mcanouil/eggla @.> Cc: Kimberley Burrows @.>; Mention @.> Subject: Re: [mcanouil/eggla] Ensure GWAS output is consistent with analysis plan (Issue #48)

I'll release eggla v0.13.0 soon, I am waiting for some continuous integration workflows/tests to complete first.

— Reply to this email directly, view it on GitHubhttps://github.com/mcanouil/eggla/issues/48#issuecomment-1247949154, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ACNZZ2KLQUM3J6LZVXARNQLV6L7QBANCNFSM6AAAAAAQBGJQSQ. You are receiving this because you were mentioned.Message ID: @.**@.>>

mcanouil commented 2 years ago

Already released (6 days ago) 😉