Machine Learning-Based Co-Expression Network Analysis Unravels Potential Fertility-Related Genes in Beef Cows
Reproductive failure is still a challenge for beef producers and a significant cause of economic loss. The increased availability of transcriptomic data has shed light on the mechanisms modulating pregnancy success. Furthermore, new analytical tools, such as machine learning (ML), provide opportunit...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-10-01
|
| Series: | Animals |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-2615/12/19/2715 |
| _version_ | 1797480936333377536 |
|---|---|
| author | Wellison J. S. Diniz Priyanka Banerjee Soren P. Rodning Paul W. Dyce |
| author_facet | Wellison J. S. Diniz Priyanka Banerjee Soren P. Rodning Paul W. Dyce |
| author_sort | Wellison J. S. Diniz |
| collection | DOAJ |
| description | Reproductive failure is still a challenge for beef producers and a significant cause of economic loss. The increased availability of transcriptomic data has shed light on the mechanisms modulating pregnancy success. Furthermore, new analytical tools, such as machine learning (ML), provide opportunities for data mining and uncovering new biological events that explain or predict reproductive outcomes. Herein, we identified potential biomarkers underlying pregnancy status and fertility-related networks by integrating gene expression profiles through ML and gene network modeling. We used public transcriptomic data from uterine luminal epithelial cells of cows retrospectively classified as pregnant (P, n = 25) and non-pregnant (NP, n = 18). First, we used a feature selection function from BioDiscML and identified <i>SERPINE3</i>, <i>PDCD1</i>, <i>FNDC1</i>, <i>MRTFA</i>, <i>ARHGEF7, MEF2B</i>, <i>NAA16,</i> ENSBTAG00000019474, and ENSBTAG00000054585 as candidate biomarker predictors of pregnancy status. Then, based on co-expression networks, we identified seven genes significantly rewired (gaining or losing connections) between the P and NP networks. These biomarkers were co-expressed with genes critical for uterine receptivity, including endometrial tissue remodeling, focal adhesion, and embryo development. We provided insights into the regulatory networks of fertility-related processes and demonstrated the potential of combining different analytical tools to prioritize candidate genes. |
| first_indexed | 2024-03-09T22:07:20Z |
| format | Article |
| id | doaj.art-50095e5fdcbd404584723d4fd053cb7f |
| institution | Directory Open Access Journal |
| issn | 2076-2615 |
| language | English |
| last_indexed | 2024-03-09T22:07:20Z |
| publishDate | 2022-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Animals |
| spelling | doaj.art-50095e5fdcbd404584723d4fd053cb7f2023-11-23T19:38:37ZengMDPI AGAnimals2076-26152022-10-011219271510.3390/ani12192715Machine Learning-Based Co-Expression Network Analysis Unravels Potential Fertility-Related Genes in Beef CowsWellison J. S. Diniz0Priyanka Banerjee1Soren P. Rodning2Paul W. Dyce3Department of Animal Sciences, Auburn University, Auburn, AL 36849, USADepartment of Animal Sciences, Auburn University, Auburn, AL 36849, USADepartment of Animal Sciences, Auburn University, Auburn, AL 36849, USADepartment of Animal Sciences, Auburn University, Auburn, AL 36849, USAReproductive failure is still a challenge for beef producers and a significant cause of economic loss. The increased availability of transcriptomic data has shed light on the mechanisms modulating pregnancy success. Furthermore, new analytical tools, such as machine learning (ML), provide opportunities for data mining and uncovering new biological events that explain or predict reproductive outcomes. Herein, we identified potential biomarkers underlying pregnancy status and fertility-related networks by integrating gene expression profiles through ML and gene network modeling. We used public transcriptomic data from uterine luminal epithelial cells of cows retrospectively classified as pregnant (P, n = 25) and non-pregnant (NP, n = 18). First, we used a feature selection function from BioDiscML and identified <i>SERPINE3</i>, <i>PDCD1</i>, <i>FNDC1</i>, <i>MRTFA</i>, <i>ARHGEF7, MEF2B</i>, <i>NAA16,</i> ENSBTAG00000019474, and ENSBTAG00000054585 as candidate biomarker predictors of pregnancy status. Then, based on co-expression networks, we identified seven genes significantly rewired (gaining or losing connections) between the P and NP networks. These biomarkers were co-expressed with genes critical for uterine receptivity, including endometrial tissue remodeling, focal adhesion, and embryo development. We provided insights into the regulatory networks of fertility-related processes and demonstrated the potential of combining different analytical tools to prioritize candidate genes.https://www.mdpi.com/2076-2615/12/19/2715biomarkercow fertilitydata miningmachine learningtranscriptomics |
| spellingShingle | Wellison J. S. Diniz Priyanka Banerjee Soren P. Rodning Paul W. Dyce Machine Learning-Based Co-Expression Network Analysis Unravels Potential Fertility-Related Genes in Beef Cows Animals biomarker cow fertility data mining machine learning transcriptomics |
| title | Machine Learning-Based Co-Expression Network Analysis Unravels Potential Fertility-Related Genes in Beef Cows |
| title_full | Machine Learning-Based Co-Expression Network Analysis Unravels Potential Fertility-Related Genes in Beef Cows |
| title_fullStr | Machine Learning-Based Co-Expression Network Analysis Unravels Potential Fertility-Related Genes in Beef Cows |
| title_full_unstemmed | Machine Learning-Based Co-Expression Network Analysis Unravels Potential Fertility-Related Genes in Beef Cows |
| title_short | Machine Learning-Based Co-Expression Network Analysis Unravels Potential Fertility-Related Genes in Beef Cows |
| title_sort | machine learning based co expression network analysis unravels potential fertility related genes in beef cows |
| topic | biomarker cow fertility data mining machine learning transcriptomics |
| url | https://www.mdpi.com/2076-2615/12/19/2715 |
| work_keys_str_mv | AT wellisonjsdiniz machinelearningbasedcoexpressionnetworkanalysisunravelspotentialfertilityrelatedgenesinbeefcows AT priyankabanerjee machinelearningbasedcoexpressionnetworkanalysisunravelspotentialfertilityrelatedgenesinbeefcows AT sorenprodning machinelearningbasedcoexpressionnetworkanalysisunravelspotentialfertilityrelatedgenesinbeefcows AT paulwdyce machinelearningbasedcoexpressionnetworkanalysisunravelspotentialfertilityrelatedgenesinbeefcows |