Scientific Workflow Interactions: An Application to Cancer Gene Identification

• Diogo Munaro Vieira Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio) https://orcid.org/0000-0002-8401-8843
• Alexandre Heine Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio) https://orcid.org/0000-0001-9235-4153
• Elvismary Molina de Armas Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio) https://orcid.org/0000-0002-0606-5994
• Cristóvão Antunes de Lanna Instituto Nacional de Câncer (INCA) https://orcid.org/0000-0002-5946-4358
• Mariana Boroni Instituto Nacional de Câncer (INCA) https://orcid.org/0000-0003-3495-9917
• Sérgio Lifschitz Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio)

Resumo

Reproducibility, resilience, and large-scale data processing have become fundamental for developing scientific research, particularly in bioinformatics. One may consider the use of Scientific Workflow Management Systems (SWfMS) to address these topics. However, user interactivity during the execution of workflows, especially with a preliminary result generated by an inner workflow task, is still an issue. We present in this paper an architecture that meets the interactive requirements of these systems, allowing the development of a flexible layer for end users to interact directly with SWfMS. Besides presenting our software solution, we show an application in the context of cancer gene identification for drug design.

Palavras-chave: Bioinformatics, Workflows, Human-in-the-loop, Interactions, Cancer