I have a number of samples where Log-likelihood in the params.txt gets set to "NA". Have you seen this before? I am attaching the command run and output text for one of the examples where it gets set to NA. Have you seen this before?
ichor_loglik_na_output_cmd.txt
ichor_loglik_na_output.txt
|
write.table(paste0("Log-likelihood:\t", format(as.numeric(hmmResults$results$loglik[i, "loglik"]), digits=4)), file = fc, col.names = FALSE, row.names = FALSE, quote = FALSE, sep = "\t") |
A couple examples that get "NA" (command and output attached for the first sample):
Gender: unknown
Tumor Fraction: 0.02015
Ploidy: 2.01
Subclone Fraction: NA
Fraction Genome Subclonal: 0
Fraction CNA Subclonal: 0
Coverage: NA
ChrY coverage fraction: NA
Likelihood model: t
Student's t mean: -0.01, -1e-04, 0.0099, 0.02, 0.03
Student's t precision: 83, 99, 91, 100, 1.4
Gamma Rate Init: 0.049
GC-Map correction MAD: 0.1919
Log-likelihood: NA
n_0 phi_0 n_est phi_est Frac_genome_subclonal Frac_CNA_subclonal loglik
0.2 2 1 1.987 0 0 667.9
0.3 2 1 1.985 0 0 667.9
0.4 2 0.98 1.997 0 0 683
0.5 2 0.98 2.01 0 0 683.1
0.6 2 0.98 2.011 0 0 683.1
0.7 2 0.98 2.018 0 0 683.1
0.8 2 0.98 2.009 0 0 683.1
0.9 2 0.98 2.02 0 0 683.1
0.95 2 0.98 2.026 0 0 683.1
Gender: unknown
Tumor Fraction: 0.02256
Ploidy: 1.178
Subclone Fraction: NA
Fraction Genome Subclonal: 0
Fraction CNA Subclonal: 0
Coverage: NA
ChrY coverage fraction: NA
Likelihood model: t
Student's t mean: -0.002, 0.0093, 0.021, 0.032, 0.043
Student's t precision: 150, 1000, 30, 29, 1.9
Gamma Rate Init: 0.035
GC-Map correction MAD: 0.03848
Log-likelihood: NA
n_0 phi_0 n_est phi_est Frac_genome_subclonal Frac_CNA_subclonal loglik
0.2 2 1 1.977 0 0 2331
0.3 2 1 1.974 0 0 2331
0.4 2 1 1.97 0 0 2331
0.5 2 1 1.964 0 0 2331
0.6 2 1 1.955 0 0 2331
0.7 2 1 1.939 0 0 2331
0.8 2 0.98 1.178 0 0 2336
0.9 2 0.99 0.8699 0 0 2334
0.95 2 0.99 0.8418 0 0 2333
And a couple with ties that don't get NA:
Gender: unknown
Tumor Fraction: 0.9486
Ploidy: 1.992
Subclone Fraction: NA
Fraction Genome Subclonal: 0
Fraction CNA Subclonal: 0
Coverage: NA
ChrY coverage fraction: NA
Likelihood model: t
Student's t mean: -0.64, 0.004, 0.39, 0.67, 0.89
Student's t precision: 110, 1100, 290, 24, 1.6
Gamma Rate Init: 0.042
GC-Map correction MAD: 0.04502
Log-likelihood: 2298
n_0 phi_0 n_est phi_est Frac_genome_subclonal Frac_CNA_subclonal loglik
0.2 2 0.051 1.992 0 0 2312
0.3 2 0.052 1.992 0 0 2312
0.4 2 0.53 1.983 0 0 2298
0.5 2 0.53 1.983 0 0 2298
0.6 2 0.53 1.983 0 0 2298
0.7 2 0.53 1.983 0 0 2298
0.8 2 0.53 1.983 0 0 2298
0.9 2 0.68 1.975 0 0 2251
0.95 2 0.68 1.975 0 0 2251
Gender: unknown
Tumor Fraction: 0
Ploidy: 1.976
Subclone Fraction: NA
Fraction Genome Subclonal: 0
Fraction CNA Subclonal: NaN
Coverage: NA
ChrY coverage fraction: NA
Likelihood model: t
Student's t mean: -0.53, 0.01, 0.36, 0.62, 0.82
Student's t precision: 28, 960, 28, 28, 1.9
Gamma Rate Init: 0.035
GC-Map correction MAD: 0.05052
Log-likelihood: 2271
n_0 phi_0 n_est phi_est Frac_genome_subclonal Frac_CNA_subclonal loglik
0.2 2 1 1.976 0 0 2271
0.3 2 1 1.973 0 0 2271
0.4 2 1 1.968 0 0 2271
0.5 2 1 1.962 0 0 2271
0.6 2 1 1.952 0 0 2271
0.7 2 1 1.936 0 0 2271
0.8 2 1 1.904 0 0 2271
0.9 2 1 1.807 0 0 2271
0.95 2 1 1.615 0 0 2271
I have a number of samples where Log-likelihood in the params.txt gets set to "NA". Have you seen this before? I am attaching the command run and output text for one of the examples where it gets set to NA. Have you seen this before?
ichor_loglik_na_output_cmd.txt
ichor_loglik_na_output.txt
ichorCNA/R/output.R
Line 132 in 11b58b3
A couple examples that get "NA" (command and output attached for the first sample):
And a couple with ties that don't get NA: