---

ClusterTAD : An unsupervised machine learning approach to detecting topologically associated domains of chromosomes from Hi-C data

---

Bioinformatics, Data Mining, Machine Learning (BDM) Laboratory,

University of Missouri, Columbia MO 65211

---

Developer:
              Oluwatosin Oluwadare
              Department of Computer Science
              University of Missouri, Columbia
              Email: [email protected]

Contact:
              Jianlin Cheng, PhD
              Department of Computer Science
              University of Missouri, Columbia
              Email: [email protected]

---

1. Content of folders:

• executable: latest ClusterTAD.jar version can be downloaded from the release tab
• examples: contains example data and outputs generated from ClusterTAD for these datasets
• src: ClusterTAD Java and MATLAB source codes
• TADs: contains identified topological domains for two mESC and mouse cortex cell type using ClusterTAD

2. Hi-C Data used in this study:

In our study, we used the normalized Hi-C matrix processed by Bing Ren's Lab in University of Calfornia, San Diego. Download the normalized Matrix here : http://chromosome.sdsc.edu/mouse/hi-c/download.html

3. Input matrix file format:

The input to ClusterTAD is a tab seperated N by N intra-chromosomal contact matrix derived from Hi-C data, where N is the number of equal-sized regions of a chromosome.

4. Usage:

4.1. Java:
To run the tool, open command line interface and type: java -jar ClusterTAD.jar Input_Matrix_file Matrix_Resolution

Parameters are as follow:

• Input_Matrix_file : A tab seperated N by N intra-chromosomal Hi-C contact matrix.
• Matrix_Resolution : Contact Matrix Resolution.

4.2. MATLAB:
Instructions on how to run the MATLAB source code is given here /src/MATLAB source code/

5. Output

ClusterTAD produces 2 folders in Output folder:

5.1. Clusters:

• Contains a .txt file that contains the cluster assignment for the diagonal for all the K values considered

5.2. TADs:

• Contains the .txt files listing the TADs extracted from each clustering and reclustering done.
• Contains the Best TAD identified based on the Quality score, labeled as "BestTAD_[nameofinputfile]_K=.txt".
• Contains a .txt file which contains a list of the extracted TAD Quality scores, file name = [nameofinputfile]_TAD_QualityScore_List.

6. Disclaimer

The executable software and the source code of ClusterTAD is distributed free of charge as it is to any non-commercial users. The authors hold no liabilities to the performance of the program.

7. Citations

Oluwadare, Oluwatosin, and Jianlin Cheng. "ClusterTAD: an unsupervised machine learning approach to detecting topologically associated domains of chromosomes from Hi-C data." BMC bioinformatics 18.1 (2017): 480.

8. Common questions

8.1. What is the format of the domain output genererated?

The domain extracted in ClusterTAD are presented in the format from.id from.cord to.id to.cord where:

• from.id : start bin id for a domain.
• from.cord : coordinate of the start bin id for a domain based on data Resolution
• to.id : end bin id for a domain.
• to.cord : coordinate of the end bin id for a domain based on data Resolution

---