Skip to main content

Table 5 Performances of the Least Absolute Shrinkage and Selection Operator (LASSO), Partial Least Square (PLS) and Sparse Partial Least Squares (SPLS), Bayes A, Bayes B and Genomic Best Linear Unbiased Prediction (GBLUP) methods to predict phenotypes measure in field trials in 2003 and 2007, with or without taking into account the genetic structures revealed by DAPC and INSTRUCT

From: Genetic diversity and trait genomic prediction in a pea diversity panel

  Thousand seed weight Beginning of flowering ( \( \sum T{}^{\circ}C \) ) Seed number
  2003 2007 2003 2007 2003 2007
MSEP       
LASSO 1931 (360) 2185 (398) 16074 (2469) 10953 (1511) 7722 (1035) 1047 (180)
LASSO-DAPC 2334 (404) 2761 (458) 17893 (3018) 13080 (1814) 9700 (1598) 1342 (206)
LASSO-INSTRUCT 1882 (364) 2455 (511) 18019 (2623) 13013 (1805) 9566 (1629) 1310 (213)
PLS 1737 (283) 2011 (312) 14733 (2196) 9883 (1247) 7978 (1019) 1003 (163)
PLS-DAPC 2058 (296) 2414 (333) 16292 (2371) 11882 (1317) 10531 (1481) 1234 (215)
PLS-INSTRUCT 1790 (249) 2040 (300) 17357 (2346) 11947 (1430) 9786 (1410) 1226 (211)
SPLS 1947 (286) 2215 (332) 16064 (2160) 11546 (1441) 7820 (985) 1079 (188)
SPLS-DAPC 2355 (321) 2696 (399) 18929 (2493) 13115 (1725) 10476 (2216) 1337 (214)
SPLS-INSTRUCT 1906 (285) 2204 (316) 18250 (2653) 12644 (1537) 9947 (1887) 1336 (215)
Bayes A 1822 (329) 2084 (344) 14823 (2517) 10099 (1431) 7631 (1020) 1304 (421)
Bayes B 1814 (327) 2051 (350) 14722 (2471) 10206 (1474) 7660 (1002) 1328 (408)
GBLUP 1759 (288) 2017 (342) 14237 (2287) 9858 (1301) 7593 (988) 1296 (409)
R 2       
LASSO 0.78 (0.06) 0.81 (0.04) 0.56 (0.11) 0.71 (0.09) 0.65 (0.07) 0.62 (0.11)
LASSO-DAPC 0.79 (0.08) 0.88 (0.04) 0.55 (0.15) 0.70 (0.11) 0.59 (0.11) 0.52 (0.16)
LASSO-INSTRUCT 0.88 (0.06) 0.91 (0.06) 0.49 (0.15) 0.70 (0.11) 0.59 (0.12) 0.52 (0.16)
PLS 0.77 (0.01) 0.79 (0.01) 0.56 (0.02) 0.76 (0.02) 0.70 (0.02) 0.69 (0.02)
PLS-DAPC 0.79 (0.01) 0.80 (0.01) 0.50 (0.02) 0.56 (0.02) 0.68 (0.02) 0.51 (0.03)
PLS-INSTRUCT 0.83 (0.01) 0.84 (0.01) 0.47 (0.03) 0.69 (0.02) 0.61 (0.02) 0.62 (0.02)
SPLS 0.74 (0.02) 0.77 (0.02) 0.59 (0.03) 0.79 (0.02) 0.67 (0.02) 0.55 (0.03)
SPLS-DAPC 0.71 (0.02) 0.73 (0.02) 0.40 (0.03) 0.73 (0.02) 0.46 (0.03) 0.41 (0.03)
SPLS-INSTRUCT 0.88 (0.01) 0.85 (0.01) 0.41 (0.03) 0.61 (0.02) 0.57 (0.03) 0.55 (0.03)
Bayes A 0.84 (0.03) 0.87 (0.02) 0.69 (0.06) 0.81 (0.03) 0.70 (0.04) 0.62 (0.11)
Bayes B 0.83 (0.03) 0.87 (0.02) 0.68 (0.06) 0.81 (0.04) 0.70 (0.04) 0.62 (0.10)
GBLUP 0.85 (0.02) 0.88 (0.02) 0.69 (0.05) 0.82 (0.03) 0.70 (0.04) 0.62 (0.10)
Q 2       
LASSO 0.50 (0.07) 0.58 (0.06) 0.20 (0.07) 0.40 (0.71) 0.38 (0.08) 0.32 (0.07)
LASSO-DAPC 0.40 (0.10) 0.47 (0.09) 0.11 (0.12) 0.30 (0.09) 0.23 (0.10) 0.11 (0.09)
LASSO-INSTRUCT 0.52 (0.09) 0.53 (0.10) 0.11 (0.08) 0.29 (0.09) 0.25 (0.10) 0.14 (0.10)
PLS 0.55 (0.06) 0.61 (0.05) 0.27 (0.07) 0.46 (0.06) 0.35 (0.09) 0.34 (0.08)
PLS-DAPC 0.47 (0.07) 0.54 (0.06) 0.19 (0.08) 0.35 (0.07) 0.17 (0.11) 0.19 (0.11)
PLS-INSTRUCT 0.54 (0.06) 0.61 (0.05) 0.13 (0.09) 0.35 (0.08) 0.22 (0.11) 0.19 (0.11)
SPLS 0.49 (0.07) 0.58 (0.05) 0.20 (0.09) 0.36 (0.08) 0.37 (0.09) 0.29 (0.08)
SPLS-DAPC 0.39 (0.09) 0.48 (0.08) 0.06 (0.11) 0.28 (0.10) 0.17 (0.18) 0.12 (0.10)
SPLS-INSTRUCT 0.51 (0.08) 0.58 (0.06) 0.10 (0.09) 0.30 (0.09) 0.21 (0.15) 0.12 (0.12)
Bayes A 0.53 (0.06) 0.61 (0.05) 0.26 (0.07) 0.45 (0.07) 0.39 (0.07) 0.27 (0.11)
Bayes B 0.53 (0.06) 0.61 (0.05) 0.27 (0.07) 0.44 (0.07) 0.38 (0.08) 0.26 (0.10)
GBLUP 0.54 (0.06) 0.62 (0.05) 0.29 (0.070) 0.46 (0.07) 0.39 (0.08) 0.28 (0.10)
  1. The best predictions are highlighted in bold. Mean square error of prediction (MSEP) indicates expected squared Euclidian distance between predicted and observed phenotypes, R 2 expressed the proportion of variance explained by the model and Q 2 evaluates the prediction quality of the model. Standard deviations are in parentheses.