Home » Abstracts » AI and Bioinformatics » Towards a Unified Pipeline: Comparing tools for Thermal Proteome Profiling experiments

Towards a Unified Pipeline: Comparing tools for Thermal Proteome Profiling experiments

Author(s)

  • Karen Druart (Presenting Author) | Institut Pateur, Core Facility MSBio Utechs | 25-28 rue du Docteur Roux, 75015, Paris, France
  • Eléonore Bouscasse | Institut Pateur, Core Facility MSBio Utechs | 25-28 rue du Docteur Roux, 75015, Paris, France
  • Rayen Elj | Institut Pateur, Core Facility MSBio Utechs | 25-28 rue du Docteur Roux, 75015, Paris, France
  • Angélique Amo | Institut Pasteur, Epigenetic Chemical Biology | 35, rue Hélène Brion, 75012, Paris, France
  • Filipe Carvalho | Institut MICALIS, Epigénétique et microbiologie cellulaire | Domaine de Vilvert, 78352, Jouy-en-Josas, France
  • Alessandro Pagliuso | Institut MICALIS, Epigénétique et microbiologie cellulaire | Domaine de Vilvert, 78352, Jouy-en-Josas, France
  • Agathe Subtil | Institut Pasteur, Cellular biology of microbial infection | 25-28 rue du Docteur Roux, 75015, Paris, France
  • Paola Arimondo | Institut Pasteur, Epigenetics and Cell Fate | 25-28 rue du Docteur Roux, 75015, Paris, France
  • Jonathan Weitzman | Institut Pasteur, Epigenetic Chemical Biology | 35, rue Hélène Brion, 75012, Paris, France
  • Mariette Matondo | Institut Pateur, Core Facility MSBio Utechs | 25-28 rue du Docteur Roux, 75015, Paris, France

Abstract

Epigenetic modifications modulate gene expression without altering the DNA sequence and are potentially reversible, making epigenetic modifying enzymes promising drug targets against antimicrobial resistance.
Identifying the protein targets of epigenetic drugs is challenging. Thermal Proteome Profiling (TPP) is powerful approach for detecting drug-affected proteins across the proteome. Although multiple analytical methods have been developed to interpret TPP data, their comparative performance remains underexplored.
We applied three analytical pipelines to a benchmark dataset to evaluate the impact of data processing tools on protein identification. Using the TPP experiment by the Savitski lab (Franken et al., 2015), in which K562 cells treated by with Panobinostat, we re-analysed both our replicated dataset and the original raw files (Childs et al., 2004).
Our results show that the choice of search engine affects the proteins identified. This underscores the importance of careful selecting and standardization of data analysis strategies in TPP workflows to ensure robust and reliable identification of drug targets. These optimized pipelines are now being applied to high-resolution datasets acquired on the Orbitrap Astral to extend target identification to more complex samples.