Skip to main content

Table 2 Simulation results

From: An EM algorithm to improve the estimation of the probability of clonal relatedness of pairs of tumors in cancer patients

   One-step maximizationEM algorithmEM algorithm - subset
N casesTrue πScenariomean(sd)rangeN 0-1mean(sd)rangeN 0-1mean(sd)rangeN 0-1
1000.101: μ=−2;σ=1.50.127(0.126)0.010-1.0000-70.086(0.036)0.010-0.2020-00.076(0.037)0.034-0.1380-0
  2: μ=−1;σ=1.00.105(0.038)0.020-0.2340-00.099(0.033)0.020-0.2120-0    
  3: μ=0.7σ=0.30.101(0.031)0.030-0.2200-00.101(0.031)0.030-0.2200-0    
 0.251: μ=−2;σ=1.50.259(0.091)0.079-0.7290-00.214(0.051)0.077-0.3870-0    
  2: μ=−1;σ=1.00.250(0.049)0.121-0.3870-00.245(0.047)0.121-0.3770-0    
  3: μ=0.7σ=0.30.252(0.043)0.130-0.3800-00.252(0.043)0.130-0.3800-0    
 0.501: μ=−2;σ=1.50.518(0.113)0.245-0.8810-00.440(0.066)0.230-0.6210-0    
  2: μ=−1;σ=1.00.498(0.055)0.325-0.6400-00.490(0.054)0.319-0.6240-0    
  3: μ=0.7σ=0.30.498(0.049)0.350-0.6200-00.498(0.049)0.350-0.6200-0    
 0.751: μ=−2;σ=1.50.756(0.116)0.495-1.0000-310.662(0.068)0.477-0.9240-00.758(0.052)0.623-0.9240-0
  2: μ=−1;σ=1.00.747(0.050)0.616-0.8810-00.738(0.049)0.609-0.8750-0    
  3: μ=0.7σ=0.30.748(0.043)0.630-0.8500-00.748(0.043)0.630-0.8500-0    
500.101: μ=−2;σ=1.50.138(0.193)0.000-1.00019-180.083(0.049)0.000-0.26511-00.083(0.070)0.000-0.26511-0
  2: μ=−1;σ=1.00.113(0.079)0.000-1.0004-10.101(0.048)0.000-0.2723-00.038(0.056)0.000-0.1253-0
  3: μ=0.7σ=0.30.100(0.042)0.000-0.2602-00.100(0.042)0.000-0.2602-00(0.000)0.000-0.0002-0
 0.251: μ=−2;σ=1.50.270(0.145)0.043-1.0000-40.210(0.071)0.043-0.4560-00.194(0.049)0.122-0.2340-0
  2: μ=−1;σ=1.00.255(0.076)0.100-0.7140-00.245(0.064)0.101-0.4470-0    
  3: μ=0.7σ=0.30.248(0.061)0.100-0.4400-00.248(0.061)0.100-0.4400-0    
 0.501: μ=−2;σ=1.50.520(0.154)0.222-1.0000-70.441(0.091)0.212-0.8040-00.64(0.097)0.494-0.8040-0
  2: μ=−1;σ=1.00.501(0.075)0.296-0.7390-00.492(0.073)0.293-0.7130-0    
  3: μ=0.7σ=0.30.498(0.069)0.320-0.7000-00.498(0.069)0.320-0.7000-0    
 0.751: μ=−2;σ=1.50.747(0.143)0.480-1.0000-520.659(0.091)0.469-0.9330-00.783(0.072)0.650-0.9330-0
  2: μ=−1;σ=1.00.746(0.075)0.530-1.0000-20.736(0.071)0.527-0.9380-00.926(0.018)0.913-0.9380-0
  3: μ=0.7σ=0.30.746(0.060)0.600-0.9200-00.746(0.060)0.600-0.9200-0    
250.101: μ=−2;σ=1.50.128(0.197)0.000-1.000101-180.099(0.079)0.000-0.44346-00.088(0.112)0.000-0.44146-0
  2: μ=−1;σ=1.00.118(0.121)0.000-1.00046-40.103(0.063)0.000-0.36526-00.056(0.084)0.000-0.36526-0
  3: μ=0.7σ=0.30.103(0.061)0.000-0.33029-00.101(0.056)0.000-0.28022-00.018(0.045)0.000-0.22822-0
 0.251: μ=−2;σ=1.50.276(0.192)0.000-1.0006-90.216(0.103)0.039-0.5430-00.222(0.108)0.039-0.4320-0
  2: μ=−1;σ=1.00.262(0.110)0.040-1.0000-10.246(0.092)0.040-0.6180-00.176 0.176-0.1760-0
  3: μ=0.7σ=0.30.250(0.087)0.040-0.5200-00.249(0.088)0.040-0.5200-0    
 0.501: μ=−2;σ=1.50.515(0.198)0.122-1.0000-190.433(0.124)0.109-0.8780-00.622(0.145)0.384-0.8780-0
  2: μ=−1;σ=1.00.505(0.110)0.201-1.0000-10.492(0.102)0.201-0.8460-00.490 0.490-0.4900-0
  3: μ=0.7σ=0.30.500(0.096)0.200-0.7600-00.500(0.096)0.200-0.7600-0    
 0.751: μ=−2;σ=1.50.752(0.175)0.358-1.0000-820.665(0.136)0.332-1.0000-20.835(0.097)0.614-1.0000-2
  2: μ=−1;σ=1.00.752(0.098)0.489-1.0000-50.741(0.094)0.483-1.0000-10.975(0.028)0.941-1.0000-1
  3: μ=0.7σ=0.30.749(0.084)0.480-0.9600-00.749(0.084)0.480-0.9600-0    
  1. The EM algorithm – subset results present the estimates obtained with the EM algorithm for the datasets where the one-step maximization gave results on the boundary. N 0-1 shows the number of times the estimate was exactly 0 - number of times it was exactly 1