A comprehensive performance evaluation on the prediction results of existing cooperative transcription factors identification algorithms — ASN Events
  1. Schedule
  2. Genome & Transcriptome Informatics 1
  3. A comprehensive performance evaluation on the prediction results of existing cooperative transcription factors identification algorithms

A comprehensive performance evaluation on the prediction results of existing cooperative transcription factors identification algorithms (#9)

Fu-Jou Lai 1 , Yueh-Min Huang 1 , Wei-Sheng Wu 2
  1. Department of Engineering Science, National Cheng Kung University, Tainan, Taiwan
  2. Computational Systems Biology Lab, Department of Electrical Engineering, National Cheng Kung University, Tainan

Background: Eukaryotic transcriptional regulation is known to be highly connected through the networks of cooperative transcriptional regulators. Measuring the cooperativity of transcriptional regulators is helpful for understanding the biological relevance of them in regulating genes. The recent advances in computational techniques led to various predictions of significant cooperative transcription factor (TF) pairs by genome-wide analysis in yeast. As each study in the related domain utilized diverse data resources and distinctive algorithms, it possessed its own merit and claimed outperforming others. However, the claim was prone to subjectivity because the study compared with a few other studies only and just used a small set of performance indices for comparison. This motivated us to develop and propose a series of measurement approaches to generate performance indices in order to objectively evaluate the prediction performance of each study. And based on these performance indices, we conducted a comprehensive performance evaluation and comparison among these studies.

Results: We collected and compiled the predicted cooperative TF pairs (PCTFPs) from 14 existing algorithms. With 7 performance indices we proposed, the cooperativity of each TF pair in each set of PCTFPs was measured and a ranking score according to the mean cooperativity of the set was given for each individual performance index. It was seen that the ranking scores of a set of PCTFPs vary with different performance indices, implying that an algorithm used in predicting cooperative TF pairs is of strength somewhere but may be of weakness elsewhere. We finally made a comprehensive ranking for these 14 sets. The results showed that Wang J’s study obtained the best performance evaluation on the prediction of cooperative TF pairs.

Conclusions: Our study has the following features: (i) It manipulated various published datasets in modelling 7 performance indices; (ii) It compared 14 sets of PCTFPs reported in literature; (iii) It carried out objective comprehensive performance evaluation on the prediction of PCTFPs; and (iv) The performance indices we proposed can be quickly introduced in the performance measurement of the new PCTFPs in the future study and helpful for making a quick comparison with others.