Skip to main content

Table 8 Estimates of index hopping incidence through concordance between low- and high-coverage sequence data in the real and simulated datasets, expressed as percentages and as isometric log-ratios

From: Impact of index hopping and bias towards the reference allele on accuracy of genotype calls from low-coverage sequencing

  Concordance by genotype (%) Isometric log-ratios
True = 0 True = 1 True = 2 3 partsa 2 partsb
0|0 1|0 2|0 0|1 1|1 2|1 0|2 1|2 2|2 0|0 vs. 1|0, 2|0 1|1 vs. 0|1, 2|1 2|2 vs. 0|2, 1|2 0|0 vs. 1|0 2|2 vs. 1|2
Observed 98.45 1.42 0.13 24.15 52.62 23.23 0.18 1.71 98.10 4.44 0.65 4.22 3.00 2.86
Simulated
 0% 99.62 0.35 0.03 23.57 52.98 23.45 0.04 0.47 99.48 5.61 0.66 5.35 4.00 3.78
 0.1% 99.53 0.44 0.03 23.59 53.52 22.89 0.08 0.52 99.40 5.47 0.68 5.06 3.83 3.72
 0.5% 99.28 0.66 0.06 23.91 53.22 22.87 0.10 0.92 98.98 5.07 0.67 4.72 3.55 3.31
 1% 98.99 0.90 0.10 23.70 53.23 23.07 0.14 1.33 98.53 4.72 0.67 4.43 3.32 3.04
 2% 98.20 1.64 0.16 23.73 52.90 23.37 0.23 2.16 97.62 4.29 0.66 4.04 2.89 2.70
 5% 96.34 3.29 0.37 23.56 53.37 23.07 0.59 4.75 94.66 3.65 0.68 3.29 2.39 2.12
Regression
 R2 0.999 0.998 0.999 0.014 0.044 0.014 0.989 1.000 0.999 0.993 0.213 0.981 0.995 0.989
 Estimate 1.74 1.77 1.47 1.28 1.45 1.43 1.58 1.48 1.67 1.46
  1. aThe 3-part isometric log-ratios take the form \(\sqrt {\frac{2}{3}} \ln \frac{(0|0)}{{\sqrt {\left( {1|0} \right)\cdot\left( {2|0} \right)} }}\)
  2. bThe 2-part isometric log-ratios take the form \(\frac{1}{\sqrt 2 }\ln \frac{(0|0)}{{\left( {1|0} \right)}}\)