Skip to main content

Table 4 Prediction accuracies using SNPs preselection based on the results of eQTL mapping of all genes (S_eQTL_A)

From: Multi-omics-data-assisted genomic feature markers preselection improves the accuracy of genomic prediction

Model

P-value cutoffsa

Prediction accuracy (Mean ± SEb)

Startle response

Starvation resistance

Female

Male

Female

Male

GBLUPd

Allc

0.204 ± 0.021

0.181 ± 0.022

0.272 ± 0.017

0.307 ± 0.015

<  0.001

0.220 ± 0.020

0.178 ± 0.023

0.268 ± 0.017

0.296 ± 0.015

<  0.0001

0.243 ± 0.020

0.191 ± 0.023

0.278 ± 0.017

0.305 ± 0.015

<  0.00001

0.241 ± 0.020

0.215 ± 0.022

0.274 ± 0.017

0.300 ± 0.014

<  0.000001

0.208 ± 0.020

0.220 ± 0.022

0.238 ± 0.018

0.288 ± 0.016

GFBLUPe

<  0.001

0.183 ± 0.023

0.181 ± 0.022

0.265 ± 0.017

0.292 ± 0.016

<  0.0001

0.216 ± 0.021

0.178 ± 0.025

0.265 ± 0.018

0.294 ± 0.015

<  0.00001

0.226 ± 0.021

0.177 ± 0.024

0.252 ± 0.017

0.276 ± 0.017

<  0.000001

0.187 ± 0.024

0.217 ± 0.022

0.237 ± 0.02

0.272 ± 0.016

  1. aP-value cutoffs: using different p-value cutoffs to preselect SNPs from whole genome sequencing (WGS) data based on the results of expression quantitative trait loci (eQTL) mapping of all genes; bSE: Standard error, cAll: All SNPs of WGS data, dGBLUP: Genomic best linear unbiased prediction, eGFBLUP: A genomic feature best linear unbiased prediction