This model is early liver cancer detection model, developed based on automated capillary electrophoresis data of cell-free DNA.
- 0 = Non-cancer
- 1 = Liver cancer
Docker >= 28.5.1 with docker compose plugin Docker installation
git clone https://github.com/cmuteam-medcmu/celiver.git
cd celiver
docker compose run --rm celiver
bash celiver-run.sh
#!/bin/bash
compose_path="docker-compose.yaml"
output_dir="test"
docker compose -f $compose_path \
-v $output_dir:/output \
run --rm \
celiver \
--input dataset/test_dataset.csv \
--prefic celiver \
--fromraw True \
--validate dataset/true_label.csv \
--outdir /output
#!/bin/bash
compose_path="docker-compose.yaml"
output_dir="test"
docker compose -f $compose_path \
-v $output_dir:/output \
run --rm \
celiver \
-c test/config.yaml
singularity pull celiver-v1.0.0.sif docker://moonipur148/celiver:latest
git clone https://github.com/cmuteam-medcmu/celiver.git
cd celiver
#!/bin/bash
celiver_root="$PWD" # Change to your full path
singularity run --cleanenv \
--pwd /work \
-B "test":/output \
-B $celiver_root:/work \
celiver.sif \
-c /work/test/config.yaml
If you use CEliver in your research, please cite our paper:
APA: Udomruk, S. et al. (2026). Machine learning–based cfDNA fragmentation profiling using automated capillary electrophoresis for early detection of hepatocellular carcinoma. Communications Medicine, 6(1). https://www.nature.com/articles/s43856-026-01437-5
BibTeX:
@article{Udomruk2026,
title = {Machine learning–based cfDNA fragmentation profiling using automated capillary electrophoresis for early detection of hepatocellular carcinoma},
author = {Udomruk, Sasimol et al.},
journal = {Communications Medicine},
year = {2026},
volume = {6},
number = {1},
doi = {https://www.nature.com/articles/s43856-026-01437-5}
}CEliver © 2025 by Cancer Research Unit, CMUTEAM, Faculty of Medicine, Chiang Mai University is licensed under CC BY-NC 4.0