In Silico Comparative Transcriptomic Analysis Of Endemic And Epidemic Kaposi’s Sarcoma

Kaposi's sarcoma (KS) is a type of cancer that forms in the lining of blood and lymph vessels. Based on the epidemiological criteria there are four subtypes of Kaposi's sarcoma — Epidemic Kaposi’s sarcoma, Classic (Mediterranean) Kaposi Sarcoma, Endemic (African) Kaposi Sarcoma and Transplant-Related Kaposi Sarcoma. While Endemic KS is unequally distributed in broad pockets around the world, the incidence of epidemic or AIDS-related KS overshadows it, since Kaposi’s Sarcoma-associated Herpesvirus (KSHV ) seroprevalence ranges from 30-90% in HIV-positive individuals. Harini Balasubramanian, a Research Fellow at Pine Biotech, performed a comparative study on the differential gene expression between endemic and epidemic Kaposi's sarcoma disease phenotypes from skin lesion biopsies to identify potential biomarkers for early detection and therapy of KS. 

Harini Balasubramanian completed her Bachelor’s degree in Industrial Biotechnology from Anna University, Chennai, India. Fascinated by the current research in epigenetics and oncology, Harini joined the Bioinformatics Research Fellowship Program at Pine Biotech. Here, she worked under the guidance of Dr. Raghavendran Lakshminarayanan, Research Consultant and Mentor at Pine Biotech, and Mr. Elia Brodsky, CEO of Pine Biotech. During the duration of the program, she worked on the following courses and sample projects:


Coming from a biotechnology background, performing a bioinformatics research project can seem like a daunting task. Before starting her research project, Harini completed introductory courses on Course 1: Introduction to Bioinformatics and Course 2: Bytes and Molecules. Followed by the introductory courses, she went on to complete few other relevant courses related to her research topic - Course 5: Transcriptomics, Course 6: Single Cell Transcriptomics, R-Coding Course 2: Introduction to Data Science (BioML), Course 7: BioML-Machine Learning for Biomedical Data, Course 8: Epigenomics and Course 9: Designing a Bioinformatics Research Project. To apply the knowledge and skills gained so far from studying the coursework, she also completed the example project on Project 03: TCGA Liver Cancer - Precision Oncology


To view Harini’s OmicsLogic student profile and learn more about the various courses and projects she has completed, visit the link - 


The data was downloaded from NCBI GEO which is available under the GEO Accession ID: GSE147704 and consisted of the following subgroups: Endemic KS, Epidemic KS and normal skin from healthy individuals. Principal Component Analysis (PCA) was performed using the Utilities Pipeline on the T-BioInfo server. There was no explicit separation in the samples, so for better clarity, the quantile normalization and PCA modules were run again, this time for endemic and epidemic samples separately. Further, Differential Gene Expression analysis was done using the Differential Expression pipeline on the T-BioInfo server. To understand the most upregulated and downregulated pathways in both the KS subtypes, the HUMAN-GAGE module, under the utility pipeline on the T-BioInfo server was performed.  

Differential Expression Pipeline

Figure 1 - Differential Expression pipeline used for analysis


Six genes in endemic KS namely - IGKV1-39, SOX11, S100A1, PCSK6, GRAP, and KRT83, and six in epidemic KS - IGKV3D-7, ORF6, PPEF1, CEMIP, PCDH17, KRT26, and IQCA1 - were found to be differentially expressed in the disease conditions. Of these, ORF6 could be a potential biomarker for detecting coinfection of KSHV in HIV-positive individuals, apart from LANA. SOX11 is also heavily involved in propagating lesions and triggering angiogenesis, which makes it another prospective biomarker for KS. She concluded that more research on the physiological and genotypic presentation of Kaposi’s sarcoma is quintessential, for discovering more therapeutic agents that can increase both the longevity and quality of life for PLWH and KS patients.


Here is what Harini had to say about her research fellowship experience: 

“During the Research Fellowship meetings, mentors are friendly and clear all doubts. So far, the transcriptomics course is well-curated. The only suggestion is to include an introductory lesson in the transcriptomics course, for the commands used (in relation to biostatistics and data visualization) in R and Python, before jumping into code for analyses. It would be easier for novice people (in programming) like me to understand the basis of the code before running the code blocks”.


To learn more about her research project, visit the link - 


Feeling motivated to work on your own research project? Then don't miss the chance to register yourself for the Bioinformatics Research Fellowship Program


Research Fellowship Program


The Bioinformatics Research Fellowship Program is a structured program that guides students through various areas of Big Data Bioinformatics Research using practical examples. During this program, we go through several high-quality research publications and learn about applications of computational biology in projects these publications describe. This allows beginners to try computational biology techniques on public domain data, making it possible to work with large files and extract meaningful information from patient samples, animal models, cell lines, and microbiota. The program provides support and mentorship, therefore it is an intensive research program involving training tasks, access to our online sessions and a guide on the implementation of learned skills to proposed research ideas. 


To learn more about the program, please visit the link: