Detection of QTNs and QTN‐by‐Environment and QTN‐by‐QTN Interactions for Cottonseed Oil Content in Upland Cotton (<scp><i>Gossypium hirsutum</i></scp> L.)

ABSTRACT Cottonseed oil is a high‐quality edible oil. A comprehensive analysis of the genetic basis of cottonseed oil formation at the molecular level will greatly promote the molecular improvement of this trait. This study performed a genome‐wide association study involving three modules integrated into the 3VmrMLM framework for cottonseed oil content (COC). In total, 49,533 single‐nucleotide polymorphisms were screened across 152 upland cotton accessions. Twenty‐six, 10 and 26 quantitative trait nucleotides (QTNs) for COC were respectively detected by the three modules: Single_env, Multi_env and Epistasis. Eight QTNs were simultaneously detected in at least two datasets or by at least two modules, and four genes associated with the stable QTNs were simultaneously enriched in GO and KEGG analyses. Using Multi_env, five QTN‐by‐environment interactions were detected, and six genes were simultaneously enriched in GO and KEGG analyses. Using Epistasis, 10 QTN‐by‐QTN interactions (QQIs) were detected, and eight protein–protein interactions involving six QQIs and eight gene pairs were predicted by a protein–protein interaction analysis. These findings provide novel insights into the genetic basis of COC.