Data-driven AI system for learning how to run transcript assemblers.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Transcript assemblers are crucial for reconstructing RNA sequences from RNA-seq data.
• Optimizing parameters for these tools is complex and time-consuming.
• Current methods often rely on default parameters, leading to suboptimal assembly accuracy.

Purpose of the Study:

• To develop an AI-driven system, AutoTuneX, for automatic prediction of optimal transcript assembler parameters.
• To enhance the accuracy and efficiency of transcript assembly from RNA-seq samples.
• To provide a data-driven strategy for optimizing sequence analysis tools.

Main Methods:

• Developed AutoTuneX, a data-driven AI system.
• Trained AutoTuneX on existing RNA-seq samples to learn parameter knowledge.
• Transferred learned knowledge to predict optimal parameters for new, unseen RNA-seq samples.
• Evaluated AutoTuneX on 1588 human RNA-seq samples using two transcript assemblers.

Main Results:

• AutoTuneX predicted parameters that led to 98% of samples achieving more accurate transcript assembly than with default parameters.
• Significant improvements in assembly accuracy were observed, with some samples showing up to a 600% increase in AUC.
• The system demonstrated effective knowledge transfer from training to testing datasets.

Conclusions:

• AutoTuneX provides an effective AI-based solution for optimizing transcript assembler parameters.
• The system significantly improves transcript assembly accuracy in RNA-seq data analysis.
• AutoTuneX represents a novel strategy for enhancing the performance of bioinformatics tools.