I'm using the moloc package for a multi-trait colocalization study. I ended up 2 dataframes.
I know the first dataframe is the posterior probability for each hypothesis.
But I'm not sure about the interpretation of the 2nd dataframe.
Take the first row as an example:
Does coloc_ppas is the posterior probability conditional on a(Only a causal variant for trait a) is TRUE?
Or does it mean the probability of re892090 is the causal variant for trait a, regardless of b and c?
And i do notice ab in [[3]]) = ab + ab,c + abc in [[2]],
But a in [[3]] != a + a,b + a,bc + a,b,c + ac,b + ab,c + ab + ac + abc in [[1]] (so all the hypo with a)
I do appreciate it if you could help.
Bw
[[1]]
prior sumbf logBF_locus PPA
a 1.00e-04 31.660903 24.6416066 2.874983e-10
a,b 1.00e-08 57.425371 43.3867777 4.446328e-03
a,c 1.00e-08 38.321883 24.2832893 2.246270e-11
a,bc 1.00e-09 61.484492 47.4458991 2.575464e-02
a,b,c 1.00e-12 64.154969 50.1163754 3.720734e-04
b 1.00e-04 25.825551 18.8062542 8.401831e-13
b,c 1.00e-08 32.422334 18.3837407 6.156310e-14
ac,b 1.00e-09 59.263052 45.2244586 2.793162e-03
c 1.00e-04 6.670301 -0.3489959 4.030459e-21
ab,c 1.00e-09 61.300269 47.2616753 2.142141e-02
ab 1.00e-05 54.660546 47.6412496 2.800614e-01
ac 1.00e-05 33.651108 26.6318110 2.103634e-10
bc 1.00e-05 29.849088 22.8297911 4.696495e-12
abc 1.00e-06 57.828136 50.8088396 6.651509e-01
zero 9.99e-01 0.000000 0.0000000 5.110692e-20
[[2]]
1118
[[3]]
coloc_ppas best.snp.coloc
a 0.9694270 rs892090
b 0.9923884 rs892090
ab 0.9666338 rs892090
c 0.6936987 rs892090
ac 0.6679441 rs892090
bc 0.6909056 rs892090
abc 0.6651509 rs892090
@clagiamba
Hi clagiamba,
I'm using the moloc package for a multi-trait colocalization study. I ended up 2 dataframes. I know the first dataframe is the posterior probability for each hypothesis. But I'm not sure about the interpretation of the 2nd dataframe. Take the first row as an example: Does coloc_ppas is the posterior probability conditional on a(Only a causal variant for trait a) is TRUE? Or does it mean the probability of re892090 is the causal variant for trait a, regardless of b and c? And i do notice ab in [[3]]) = ab + ab,c + abc in [[2]], But a in [[3]] != a + a,b + a,bc + a,b,c + ac,b + ab,c + ab + ac + abc in [[1]] (so all the hypo with a)
I do appreciate it if you could help. Bw
[[1]] prior sumbf logBF_locus PPA a 1.00e-04 31.660903 24.6416066 2.874983e-10 a,b 1.00e-08 57.425371 43.3867777 4.446328e-03 a,c 1.00e-08 38.321883 24.2832893 2.246270e-11 a,bc 1.00e-09 61.484492 47.4458991 2.575464e-02 a,b,c 1.00e-12 64.154969 50.1163754 3.720734e-04 b 1.00e-04 25.825551 18.8062542 8.401831e-13 b,c 1.00e-08 32.422334 18.3837407 6.156310e-14 ac,b 1.00e-09 59.263052 45.2244586 2.793162e-03 c 1.00e-04 6.670301 -0.3489959 4.030459e-21 ab,c 1.00e-09 61.300269 47.2616753 2.142141e-02 ab 1.00e-05 54.660546 47.6412496 2.800614e-01 ac 1.00e-05 33.651108 26.6318110 2.103634e-10 bc 1.00e-05 29.849088 22.8297911 4.696495e-12 abc 1.00e-06 57.828136 50.8088396 6.651509e-01 zero 9.99e-01 0.000000 0.0000000 5.110692e-20 [[2]] 1118 [[3]] coloc_ppas best.snp.coloc a 0.9694270 rs892090 b 0.9923884 rs892090 ab 0.9666338 rs892090 c 0.6936987 rs892090 ac 0.6679441 rs892090 bc 0.6909056 rs892090 abc 0.6651509 rs892090