Projects Overview

CovRadar

CovRadar is a tool for genomic surveillance of the coronavirus spike protein. The spike protein contains the receptor binding domain that is used as a target for most vaccine candidates. CovRadar consists of a workflow pipeline and a web application that enable the analysis and visualization of hundreds of thousands of sequences.

[Project Description] [Project Website] [Code]

Towards Generalizable Hierarchical Classification with HiClass

HiClass is a Python library for local hierarchical classification fully compatible with scikit-learn. It provides generalizable implementations of the most popular machine learning models for local hierarchical classification. Furthermore, the library includes metrics to evaluate model performance on hierarchical data.

[Project description] [Code] [Documentation] [Manuscript]

Using machine learning and multi-layer molecular networks to predict drug response

Due to the complexity of biological systems, finding the right treatment for a specific disease can be challenging. We derive computational models based on networks and machine learning that can deal with multi-layered data and thus help our understanding of the effects of drugs.

[Project description]

Real-time sequence-based characterization of pathogens

DNA sequencing has shown to be an excellent tool for pathogen detection and pathogen characterization. In particular, the time to detection plays a vital role in clinical diagnostics. Therefore, we are developing various real-time tools to analyze DNA sequencing data to decrease the time to report the detected pathogen (and antibiotic resistance gene), which can fasten diagnosis and accelerate the prescription of the correct medical treatment.

[Project Description]

DAKI-FWS: AI-based early warning system for the German economy

In the DAKI-FWS project, we aim to develop an AI-based analysis platform to generate forecasts for extreme events such as pandemics. We at HPI focus on data-driven analysis and development of models using many different data sources, e.g., epidemiological, medical, sequence, demographic, contact, and mobility data for pandemic modeling and forecasting.

[Project Description] [Project Website]

Contact: anna-juliane.schmachtenberg(at)hpi.de

Interpretable deep learning for novel pathogen detection from DNA sequences

Regular emergence of novel pathogens is one of the greatest threats to global health, and synthetic DNA must be screened for potential threats. However, standard approaches for pathogen detection can only recognize agents that are already known. We solve this problem by training deep neural networks that predict if a DNA read originates from potential human pathogens (bacteria, viruses or fungi) or viruses and microbes that do not infect humans.

[Project description] [Code]

Collaboration projects with the Hasso Plattner Institute for Digital Health at Mount Sinai (HPI･MS) on Electronic Health Records

Collaboration projects with our partner institution, the Hasso Plattner Institute for Digital Health at Mount Sinai (HPI･MS) at the Icahn School of Medicine at Mount Sinai, New York.

[Project descriptions]