Chromatin accessibility · Research internship
Bulk ATAC-seq analysis of mature and memory NK cells
Do mature and memory human NK cells show reproducible differences in chromatin accessibility across public bulk ATAC-seq samples?
Public-data analysis completed at UMG as the internship component of Göttingen’s M.CoBi.310 Systems Biology course. The public cohort has six donors; four have both cell states, while two have memory samples only.
View GitHub repositoryATAC-seq / workflow schematic
- 01Quality control
- 02Alignment
- 03Peak calling
- 04Region counts
- 05Donor-aware testing
- 06Interpretation
I designed and implemented the analysis workflow and its public documentation for 10 public samples, from quality control and alignment through donor-aware differential testing, motif exploration, genome-browser review, documentation, and a synthetic demonstration.
Completed during a bioinformatics internship at University Medical Center Göttingen using public data.
- 01FastQC quality review and Bowtie2 alignment
- 02SAMtools processing and MACS3 peak calling
- 03Consensus-region counting with featureCounts
- 04Donor-aware differential accessibility with DESeq2
- 05HOMER motif exploration and IGV review
- 59,186 of 112,759 consensus regions passed the count filter (at least 10 counts in at least five samples).
- The workflow connects quality review, alignment, consensus-region counting, donor-aware differential testing, motif exploration, and genome-browser review.
- The repository preserves the complete workflow, statistical results, tutorial material, and a synthetic demo for reproducibility.
- The public cohort was small and only four donors had complete mature–memory pairs, limiting statistical power.
- Exploratory motif and browser observations are not causal or mechanistic evidence.
- Public-data analyses remain sensitive to cohort composition, preprocessing choices, and available metadata.
- Methods and cohort design (opens in a new tab)
Sample structure, filtering, and donor-aware design.
- Reproducibility and data availability (opens in a new tab)
The public release is a curated snapshot, not an end-to-end rerun of protected internship outputs.
- Tutorial and synthetic demo (opens in a new tab)
A lightweight demonstration; it does not reproduce the public-cohort results.