HCSeer employs the Kernel Density Estimation (KDE) algorithm and the Expectation-Maximization (EM) algorithm from machine learning to calculate the variant density on the amino acid sequence of each gene and identify potential hot and cold spot regions.

1. numpy (v>=1.26.4)
2. pandas (v>=2.2.2)
3. scikit-learn (v>=1.4.2)
4. scipy (v>=1.13.1)
5. python (v>=3.6)

These scripts do not require compilation.

How to use Auto-annotation.py scripts to annotate whether our variants are located in cold and hot spot regions

python Auto-annotation.py -i /path/to/your/input.vcf -o /path/to/your/output.vcf -buildver hg38

• 00_database-processing.sh

  Preprocess the data in the Clinvar database

• 01_clinvar-annotation.sh

  Annotate preprocessed data with an annotator

• 02_annotation-data-processing.py

  Process the data generated by annotations

• 03_main.py

  Kernel density estimation algorithm

• 04_EM.py

  Maximum Expectation Algorithm

• 05_Initial-result-processing.py

  Process the initial result to obtain the final result

• 06_compute_p-value.py

  Calculate p-value

• 07_profile-coefficient.py

  Calculate contour coefficient

• hg19-to-hg38.sh

  Script for converting hg19 genome version coordinates to hg38 version

• inAutoPVS1.py

  Compared to AutoPVS1

• PP2_BP1_PM1_PS1.py

  Combining PP2, PM5, PS1, BP1

• process_function.py

  Some file processing functions

• result-analysis.py

  Result analysis

If you use EpiAgent in your research, please cite our paper:

Xia, X., Zhao, G., Yuan, X. (2026). HCSeer: A Classification Tool for Human Genetic Variant Hot and Cold Spots Designed for PM1 and Benign Criteria in the ACMG Guideline. In: Tang, J., Lai, X., Cai, Z., Peng, W., Wei, Y. (eds) Bioinformatics Research and Applications. ISBRA 2025. Lecture Notes in Computer Science(), vol 15756. Springer, Singapore. https://doi.org/10.1007/978-981-95-0698-9_1