Skip to main content

Table 3 The computing performance tests of LMM, BSLMM, BayesR, and KAML methods by using the WTCCC1 dataset

From: KAML: improving genomic prediction accuracy of complex traits using machine learning determined parameters

Traits/time (hours)

Methods

LMM

BSLMM

BayesR

KAML

CAD

0.01 (0.00)

13.78 (2.55)

44.93 (2.45)

0.30 (0.01)

HT

0.01 (0.00)

27.15 (5.25)

42.83 (2.46)

0.32 (0.02)

T2D

0.01 (0.00)

38.11 (6.15)

45.85 (2.17)

0.32 (0.01)

BD

0.01 (0.00)

23.5 (6.26)

40.21 (2.26)

0.35 (0.03)

CD

0.01 (0.00)

56.55 (5.02)

32.87 (0.80)

0.35 (0.01)

RA

0.01 (0.00)

3.85 (0.27)

33.22 (0.78)

0.40 (0.00)

T1D

0.01 (0.00)

5.98 (0.27)

34.23 (1.02)

0.4 (0.00)

Average

0.01

24.13

39.16

0.35

  1. Computing performance tests were conducted in a Red Hat Enterprise Linux sever with 2.20 GHz Intel(R) Xeon(R) 132 CPUs E7-8880 v4, and 2 TB memory. The computing time records and the standard deviations are described in Table 2. The computing performances of BSLMM and BayesR methods were tested using their default settings
Back to article page

Genome Biology

ISSN: 1474-760X

Contact us