Peax - Pattern Similarity Expriments

We compared the perceived similarity of patterns in sequential data at 3 different sizes. The sequential data is a DNase-seq experiment from the ENCODE project. We extracted 9 different patterns at increasing sizes of 3 kilobase pairs (kb), 12 kb, and 120 kb.

Experiment 1: groupwise comparison of 7 similarity search techniques

Experiment 2: pairwise comparison of CAE and ED

Reproduce results

conda env create --file environment.yml
jupyter lab

The staistical analysis and the code for generating the figures is available in the following three notebooks:

• user-study-1-results.ipynb
• user-study-2-results.ipynb
• user-study-1-2-combined-results.ipynb