🎗️Gene expression study of breast cancer using Welch Satterthwaite t-test, Kaplan-Meier estimator plot and Huber loss robust regression model
Published our research on the Gene expression study of breast cancer using Welch Satterthwaite t-test, Kaplan-Meier estimator plot and Huber loss robust regression model. Our findings suggest that the Huber loss robust regression model is an effective algorithm for modeling the relationship between control and breast tumor samples. This algorithm produced a correlation coefficient of 0.4398 and a mean absolute error of 1.069 ± 0.020, indicating a good fit between the model and the data.
Published in Journal of King Saud University — Science and the publishing services is provided by Elsevier with Impact Factor 3.829
Read our paper here: link
Objective
Breast Cancer (BC) is one of the deadliest diseases in women, causing thousands of deaths annually despite the advent of high-throughput genomic platforms in the recent past. Microarray based gene expression profiling with different statistical methods have been extensively used to under stand the disease at the molecular level. We plan to apply Welch Satterthwaite t-test, Kaplan-Meier estimator plot and Huber Loss robust regression model on microarray data to improve the analysis and find biomarkers for future diagnosis, prognosis, and treatment.
Methods
We retrieved microarray data (GSE10810 dataset) of 31 breast tumor samples and 27 normal breast samples from Gene Expression Omnibus (GEO, NCBI). Welch Satterthwaite t-test was applied to identify the most statistically significant genes, Huber loss robust regression model was applied to investigate the existing mathematical relations between tumor and control variables, and Kaplan-Meier Plotter was used to confirm their association with overall metastatic relapse-free survival of BC patients.
Results
We identified 1837 differentially expressed genes, including 638 overexpressed (COL11A1, KIAA0101, S100P, GJB2, TOP2A, LINC01614, RRM2, INHBA, C15orf48 and CKS2) and 1199 under expressed (LEP, ADIPOQ, PLIN1, PCK1, PCOLCE2, ADH1B, LYVE1, FABP4, ABCA8, and CHRDL1) genes passing the thresh old (fold change ± 2 and p value < 0.001). KM analysis revealed 12 out of 20 DEGs (log rank p value < 0.05)
as potential prognostic and therapeutic biomarkers.
Conclusion
Huber loss robust regression model was found to be an effective algorithm for the mathematical relationship between the control and breast tumor samples with co-relation coefficient of 0.4398 and mean absolute error of 1.069 ± 0.020. In conclusion, with high mathematical confidence, we detected DEGs have high potential to be BC biomarkers using Welch t-test and Kaplan-Meier plot having minimum underlying assumptions.
Many thanks for reading this post!🙏.
If you found this content helpful😊, please LIKE 👍, SHARE, and FOLLOW to stay updated on our future posts.
If you have a moment, I encourage you to see my other posts below:
