Implementation of data structures such as Q-Gram index, suffix array, Burrows-Wheeler transformation, and push-down automata, along with algorithms for exact string matching (Horspool), global alignments (Needleman-Wunsch), and BLAST neighborhood generation.
Practicum: German Heart Center at the Charité (DHZC).
Integration and analysis of transcriptomics, (phospho-)proteomics, and metabolomics data using advanced bioinformatics tools in R and Python to identify and understand key molecular pathways, networks, and interactions.
Web application for visualizing bicycle thefts in Berlin with filter functions and interactive plots based on a locally created PostgreSQL database.
Backend: Neural Network model trained using PyTorch to predict the risk of diabetes based on patient data, on features like pregnancies, glucose levels, blood pressure, and more.
Practicum: Berlin Institute of Health (BIH@Charité).
Developing ML models to diagnose “silent” heart attacks from ECG waveforms using Logistic Regression, Support Vector Machines, Convolutional & Recurrent Neural Networks.
Web App for constructing phylogenetic trees using various clustering methods and algorithms
like Hierarchical Clustering: Hierarchical clustering, the Neighbor-Joining algorithm and Maximum Parsimony.
[bash, fastQC, samtools, bedtools, vcftools, bwa, gatk || Nextflow ]
[R | Seurat, edgeR, DESeq2 ]
[Python | Scanpy, scviTools ]
[Julia | SingleCellGenomics.jl ]