Skip to main content

Table 3 Genomic relationship matrix calculation times (sec)

From: Second-generation PLINK: rising to the challenge of larger and richer datasets

Dataset Machine GCTA PLINK 1.90 Ratio
synth1p Mac-2 222.2 7.2 31
  Mac-12 184.7 5.0 37
  Linux32-8 248.4 10.9 22.8
  Linux64-512 4.4 9.6 0.46
  Win32-2 373.1 39.3 9.5
  Win64-2 367.2 6.6 56
synth2p Mac-2 nomem 805.8  
  Mac-12 17.0 k 138.3 123
  Linux32-8 nomem 1153.4  
  Linux64-512 65.1 318.9 0.20
  Win32-2 nomem 2007.2  
  Win64-2 nomem 450.1  
chr1snp Mac-2 nomem 87.1  
  Mac-12 2260.9 50.9 44.4
  Linux32-8 nomem 94.3  
  Linux64-512 58.3 91.6 0.64
  Win32-2 nomem 317.5  
  Win64-2 nomem 65.7  
  1. This involves a variance-normalizing distance function which cannot be efficiently computed with just bit population counts. PLINK 1.9’s lookup table-based algorithm is slower than GCTA 1.24 on heavily multicore machines (see the Linux64-512 results), but has complementary advantages in portability, accuracy, and memory efficiency.