ABS 009: Computational Pipeline Identifies Novel TCR-T Target Epitopes in Sarcoma Through Integrated RNA-Sequencing and Immunopeptidomics Analysis

Punarvash V. Mitta ¹², Yulun Chiu ², Cassian Yee ²

¹ Wiess School of Natural Sciences, Rice University
² Melanoma Medical Oncology, MD Anderson Cancer Center

The Van Wickle Journal (2026) Volume 2, ABS009

Introduction: Adoptive T cell therapies, specifically Endogenous T Cell (ETC) Therapies, have shown promise for sarcoma treatment, with objective response rates reaching up to 50% in clinical studies. Despite this potential, only two validated TCR T targets, NY ESO 1 and MAGE A4, currently exist for any sarcoma subtypes, limiting therapeutic options for a malignancy with less than 15% late stage survival. To address this gap, the objective of this project was to develop a computational pipeline capable of identifying novel, naturally presented epitopes suitable for sarcoma TCR T therapy development by integrating RNA sequencing expression data with mass spectrometry based immunopeptidomics.

Gene expression data from TCGA SARC (n=262), GTEx normal tissues (n=3270), and GEO Ewing sarcoma samples (n=46) were obtained and uniformly processed using the UCSC Xena TOIL platform. Differential expression analysis was performed using DESeq2 to identify genes overexpressed in sarcoma relative to normal tissues. Candidate targets were integrated with immunopeptidomics databases and prioritized based on expression specificity and predicted HLA binding affinity.

The pipeline identified predominant presentation of peptides derived from cell cycle progression and mitotic replication genes in sarcoma. Multiple epitopes demonstrated significant upregulation of their parent genes across sarcoma subtypes while also exhibiting strong predicted binding affinity for common HLA allotypes. Among these, LOXHD1 emerged as a particularly promising target, demonstrating specific expression in Ewing sarcoma and favorable structural binding predictions.

This integrated computational approach successfully identified and validated novel TCR T target candidates for sarcoma, establishing a scalable framework for expanding TCR T therapeutic development in sarcoma and potentially other malignancies.

Methods: Gene expression datasets from TCGA SARC, GTEx normal tissues, and GEO Ewing sarcoma cohorts were collected and uniformly processed using the UCSC Xena TOIL RNA sequencing pipeline. Differential expression analysis was performed with DESeq2 to identify genes significantly overexpressed in sarcoma relative to normal tissues. Candidate genes were cross referenced with immunopeptidomics databases to identify naturally presented MHC associated peptides. Predicted peptide HLA binding affinities were evaluated across common HLA allotypes using computational structural modeling and binding prediction algorithms. To experimentally validate candidate epitopes, MHC bound peptides were isolated from tumor lysates through immunoaffinity purification and analyzed using high resolution liquid chromatography tandem mass spectrometry (LC MS/MS). Cross validation analyses were additionally performed across multiple sarcoma subtypes to evaluate the reproducibility and broader applicability of identified targets.

Results: The computational pipeline identified multiple naturally presented epitopes derived from genes involved in cell cycle progression and mitotic replication pathways. Several candidate targets demonstrated significant overexpression in sarcoma compared to normal tissues and exhibited strong predicted binding affinity for common HLA allotypes. Cross validation across sarcoma subtypes showed consistent upregulation of multiple targets, supporting broad therapeutic applicability. LOXHD1 emerged as a particularly promising Ewing sarcoma specific target due to its restricted expression profile and favorable structural binding predictions. Experimental immunopeptidomics analysis confirmed natural HLA presentation of several prioritized epitopes, supporting the robustness of the integrated computational framework.

Discussion: This study demonstrates the utility of integrating transcriptomic and immunopeptidomic analyses to identify TCR T targets in sarcoma. By combining large scale RNA sequencing datasets with experimental validation of presented epitopes, the pipeline expands the repertoire of sarcoma TCR T targets beyond NY ESO 1 and MAGE A4. The identification of broadly expressed and subtype specific candidates, including LOXHD1, highlights the potential for pan sarcoma and precision immunotherapy approaches. Future studies will focus on validating the immunogenicity of additional epitopes and sequencing cytotoxic T lymphocyte derived T cell receptors to support development of translatable TCR T therapies.