Predicting transcription factor activity using prior biological information.

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Transcription factor (TF) dysregulation is a key driver of disease pathogenesis.
• Accurate detection of aberrant TF activity is crucial for understanding disease mechanisms.

Purpose of the Study:

• To introduce Priori, a novel computational method for predicting TF activity from RNA sequencing (RNA-seq) data.
• To highlight Priori's advantages over existing methods in sensitivity, specificity, and biological relevance.

Main Methods:

• Priori integrates literature-supported regulatory information to establish TF-target gene relationships.
• Linear models are employed to quantify the impact of TF regulation on target gene expression.
• A third-party benchmarking pipeline evaluated Priori against 11 other methods using 124 single-gene perturbation experiments.

Main Results:

• Priori demonstrated superior sensitivity and specificity in detecting aberrant TF activity compared to 11 other methods.
• Application of Priori to patient datasets revealed unique determinants of breast cancer survival.
• Priori identified key mediators of drug response in leukemia patient samples.

Conclusions:

• Priori is a highly sensitive and specific method for predicting transcription factor activity from RNA-seq data.
• Priori offers unique insights into disease pathogenesis, survival determinants, and drug response mechanisms.
• This method has significant potential for advancing precision medicine and therapeutic target discovery.