Computationally unraveling genetic factors underlying morphological patterns of wheat’s root growth: a scientific data analysis approach
- 1Department of Agronomy, National Taiwan University
- 2Department of Statistics, University of California at Davis
Abstract
A collection of 53 wheat varieties’ root-growth trajectories was clustered into seven distinct morphological growth patterns. Across more than 30K DNA variants, we used Categorical Exploratory Data Analysis (CEDA) to examine all 21 pairwise contrasts between these patterns. Short single- and multi-locus genotype combinations were evaluated by entropy reduction and a finite-sample reliability check based on alternative and null ensembles of contingency tables. Bipartite networks between wheat varieties and retained genotype combinations were displayed as presence–absence heatmaps to reveal pair-specific block structures. As an out-of-pair negative control, the genotype combinations derived from each focal pair were examined, without feature reselection, in varieties belonging to the other five phenotype groups. These excluded groups did not reproduce the focal-pair block structure, supporting the pair specificity of the displayed information within the analyzed dataset. For methodological context, the same samples, markers, and phenotypes were also analyzed with mixed-model association, interaction-detection, and machine-learning methods. Among the four recurrent CEDA markers, marker-level support was strongest for RAC875_c60191_947, which had a FarmCPU PC1 Benjamini–Hochberg-adjusted value of q=0.0275 and ranked ninth by random-forest importance in the same dataset. These comparisons characterized complementary, method-specific outputs. CEDA thereby characterized genotype-pattern information associated with MRL trajectories in this wheat panel. Across the 21 pairwise contrasts, recurring category-associated genotypes highlighted three putative positional or functional candidates at two loci.
Related
- Distinct neural oscillatory patterns underlying music-evoked emotions: Evidence from MEG and subjective reports
- The fluorescence patterns of antinuclear antibodies: morphological characteristics, immunological correlations and clinical significance
- An update on the mosquito fauna of Aguascalientes, Mexico: morphological and molecular identifications, host-feeding patterns, and the removal of Aedes idahoensis (Theobald) from Mexico
- Exploiting Low-Rank Structure in Genomic Relationship Matrices for Computationally Efficient Optimum Contribution Selection
- Computationally Optimized H1 and H3 Hemagglutinin Messenger RNA Vaccines Confer Broad Protective Immunity Against Modern Influenza Viruses
BibTeX
@article{lin2026computationally,
title = {{Computationally unraveling genetic factors underlying morphological patterns of wheat’s root growth: a scientific data analysis approach}},
author = {Shang-Chieh Lin and Hao-Chia Lo and Chih-Wei Tung and Li-Yu Liu and Fushing Hsieh},
journal = {Scientific Reports},
year = {2026},
month = sep,
doi = {10.1038/s41598-026-68718-w},
url = {https://doi.org/10.1038/s41598-026-68718-w},
note = {CC BY},
}
RIS
TY - JOUR
TI - Computationally unraveling genetic factors underlying morphological patterns of wheat’s root growth: a scientific data analysis approach
AU - Shang-Chieh Lin
AU - Hao-Chia Lo
AU - Chih-Wei Tung
AU - Li-Yu Liu
AU - Fushing Hsieh
PY - 2026
DA - 2026/09/01
JO - Scientific Reports
DO - 10.1038/s41598-026-68718-w
UR - https://doi.org/10.1038/s41598-026-68718-w
AB - A collection of 53 wheat varieties’ root-growth trajectories was clustered into seven distinct morphological growth patterns. Across more than 30K DNA variants, we used Categorical Exploratory Data Analysis (CEDA) to examine all 21 pairwise contrasts between these patterns. Short single- and multi-locus genotype combinations were evaluated by entropy reduction and a finite-sample reliability check based on alternative and null ensembles of contingency tables. Bipartite networks between wheat varieties and retained genotype combinations were displayed as presence–absence heatmaps to reveal pair-specific block structures. As an out-of-pair negative control, the genotype combinations derived from each focal pair were examined, without feature reselection, in varieties belonging to the other five phenotype groups. These excluded groups did not reproduce the focal-pair block structure, supporting the pair specificity of the displayed information within the analyzed dataset. For methodological context, the same samples, markers, and phenotypes were also analyzed with mixed-model association, interaction-detection, and machine-learning methods. Among the four recurrent CEDA markers, marker-level support was strongest for RAC875_c60191_947, which had a FarmCPU PC1 Benjamini–Hochberg-adjusted value of q=0.0275 and ranked ninth by random-forest importance in the same dataset. These comparisons characterized complementary, method-specific outputs. CEDA thereby characterized genotype-pattern information associated with MRL trajectories in this wheat panel. Across the 21 pairwise contrasts, recurring category-associated genotypes highlighted three putative positional or functional candidates at two loci.
SN - 2045-2322
ER -