A5029955333- Bioinformatics and Genomic Networks
- Gene expression and cancer classification
- Nutrition, Genetics, and Disease
- Computational Drug Discovery Methods
- Privacy-Preserving Technologies in Data
- Health, Environment, Cognitive Aging
- Molecular Biology Techniques and Applications
- Metabolomics and Mass Spectrometry Studies
- Artificial Intelligence in Healthcare and Education
- Cancer Genomics and Diagnostics
- Genetic Associations and Epidemiology
- Domain Adaptation and Few-Shot Learning
- Advanced Proteomics Techniques and Applications
- Genetics, Bioinformatics, and Biomedical Research
- Genomics and Rare Diseases
- Microbial Metabolic Engineering and Bioproduction
- Gene Regulatory Network Analysis
- Asthma and respiratory diseases
- Monoclonal and Polyclonal Antibodies Research
- Radiomics and Machine Learning in Medical Imaging
- Ubiquitin and proteasome pathways
- Protein Degradation and Inhibitors
- AI in cancer detection
- Ethics in Clinical Research
- Cell Image Analysis Techniques
Universität Hamburg
2021-2025
Technical University of Munich
2020-2024
Hamburg Institut (Germany)
2024
University Medical Center Hamburg-Eppendorf
2024
Helmholtz Zentrum München
2024
Weihenstephan-Triesdorf University of Applied Sciences
2022
Bielefeld University
2018-2021
Addgene
2017
Vavilov Institute of General Genetics
2015-2016
Lomonosov Moscow State University
2016
Historically, gene expression has been shown to be the most informative data for drug response prediction. Recent evidence suggests that integrating additional omics can improve prediction accuracy which raises question of how integrate omics. Regardless integration strategy, clinical utility and translatability are crucial. Thus, we reasoned a multi-omics approach combined with datasets would relevance.
Abstract In recent decades, the development of new drugs has become increasingly expensive and inefficient, molecular mechanisms most pharmaceuticals remain poorly understood. response, computational systems network medicine tools have emerged to identify potential drug repurposing candidates. However, these often require complex installation lack intuitive visual mining capabilities. To tackle challenges, we introduce Drugst.One, a platform that assists specialized in becoming...
Limited data access has hindered the field of precision medicine from exploring its full potential, e.g. concerning machine learning and privacy protection rules.Our study evaluates efficacy federated Random Forests (FRF) models, focusing particularly on heterogeneity within between datasets. We addressed three common challenges: (i) number parties, (ii) sizes datasets (iii) imbalanced phenotypes, evaluated five biomedical datasets.The FRF outperformed average local models performed...
Background Machine learning and artificial intelligence have shown promising results in many areas are driven by the increasing amount of available data. However, these data often distributed across different institutions cannot be easily shared owing to strict privacy regulations. Federated (FL) allows training machine models without sharing sensitive In addition, implementation is time-consuming requires advanced programming skills complex technical infrastructures. Objective Various tools...
Hypertension and bronchial asthma are a major issue for people's health. As of 2014, approximately one billion adults, or ~ 22% the world population, have had hypertension. 2011, 235–330 million people globally been affected by 250,000–345,000 died each year from disease. The development effective treatment therapies against these diseases is complicated their comorbidity features. This often problem in diagnosis treatment. Hence, this study bioinformatical methodology analysis two...
Abstract Motivation The goal of pharmacogenomics is to predict drug response in patients using their single- or multi-omics data. A major challenge that clinical data (i.e. patients) with outcome very limited, creating a need for transfer learning bridge the gap between large pre-clinical datasets (e.g. cancer cell lines), as source domain, and target domain. Two discrepancies exist datasets: (i) input space, gene expression due difference basic biology, (ii) output different measures...
Abstract Asthma and hypertension are complex diseases coinciding more frequently than expected by chance. Unraveling the mechanisms of comorbidity asthma is necessary for choosing most appropriate treatment plan patients with this comorbidity. Since both have a strong genetic component in article we aimed to find study genes simultaneously associated hypertension. We identified 330 shared found that they form six modules on interaction network. A overlap between was level eQTL regulated...
Abstract Cancer is a heterogeneous disease characterized by unregulated cell growth and promoted mutations in cancer driver genes some of which encode suitable drug targets. Since the distinct set can vary between within types, evidence-based selection drugs crucial for targeted therapy following precision medicine paradigm. However, many putative not be directly, suggesting an indirect approach that considers alternative functionally related targets gene interaction network. Once potential...
Despite the significant progress in accuracy and reliability mass spectrometry technology, as well development of strategies based on isotopic labeling or internal standards recent decades, systematic biases originating from non-biological factors remain a challenge data analysis. In addition, wide range available normalization methods renders choice suitable method challenging. We systematically evaluated 17 two batch effect correction methods, originally developed for pre-processing DNA...
Abstract Tumors of the jaws may represent different human disorders and frequently associate with pathologic bone fractures. In this report, we analyzed two affected siblings from a family Russian origin, history dental tumors jaws, in correspondence to original clinical diagnosis cementoma consistent gigantiform (GC, OMIM: 137575). Whole exome sequencing revealed heterozygous missense mutation c.1067G > A (p.Cys356Tyr) ANO5 gene these patients. To date, autosomal-dominant mutations have...
In vitro selection of antibodies from large repertoires immunoglobulin (Ig) combining sites using combinatorial libraries is a powerful tool, with great potential for generating in vivo scavengers toxins. However, addition maturation function necessary to enable these selected more closely mimic the full mammalian immune response. We approached this goal quantum mechanics/molecular mechanics (QM/MM) calculations achieve silico. preselected A17, an Ig template, naïve library its ability...
Comorbidity, a co-incidence of several disorders in an individual, is common phenomenon. Their development governed by multiple factors, including genetic variation. The current study was set up to look at associations between isolated and comorbid diseases bronchial asthma hypertension, on one hand, single nucleotide polymorphisms associated with regulation gene expression (eQTL), the other hand. A total 96 eQTL SNPs were genotyped 587 Russian individuals. Bronchial alone found be rs1927914...
Abstract Comorbid states of diseases significantly complicate diagnosis and treatment. Molecular mechanisms comorbid asthma hypertension are still poorly understood. Prioritization is a way for identifying genes involved in complex phenotypic traits. Existing methods prioritization consider genetic, expression evolutionary data, molecular-genetic networks other. In the case networks, as rule, protein-protein interactions KEGG used. ANDSystem allows reconstructing associative gene which...
The prevalence of comorbid diseases poses a major health issue for millions people worldwide and an enormous socio-economic burden society. molecular mechanisms the development comorbidities need to be investigated. For this purpose, workflow system was developed aggregate data on biomedical entities from heterogeneous sources. process integrating merging all sources implemented as semi-automatic pipeline that provides import, fusion, analysis highly connected in Neo4j database GenCoNet. As...
Abstract Motivation Historically, gene expression has been shown to be the most informative data for drug response prediction. Recent evidence suggests that integrating additional omics can improve prediction accuracy which raises question of how integrate omics. Regardless integration strategy, clinical utility and translatability are crucial. Thus, we reasoned a multi-omics approach combined with datasets would relevance. Results We propose MOLI, M ulti- O mics L ate I ntegration method...
Clinical time-to-event studies are dependent on large sample sizes, often not available at a single institution. However, this is countered by the fact that, particularly in medical field, individual institutions legally unable to share their data, as data subject strong privacy protection due its particular sensitivity. But collection, and especially aggregation into centralized datasets, also fraught with substantial legal risks outright unlawful. Existing solutions using federated...
Abstract Motivation One of the main goals precision oncology is to predict response a patient given cancer treatment based on their genomic profile. Although current models for drug prediction are becoming more accurate, they also ‘black boxes’ and cannot explain predictions, which particular importance in treatment. Many do not leverage prior biological knowledge, such as hierarchical information how proteins form complexes act together pathways. Results In this work, we use knowledge...
Abstract Motivation Identification of differentially expressed genes is necessary for unraveling disease pathogenesis. This task complicated by the fact that many diseases are heterogeneous at molecular level and samples representing distinct subtypes may demonstrate different patterns dysregulation. Biclustering methods capable identifying follow a similar expression pattern only in subset hence can consider heterogeneity. However, biologically significant reproducible sets remain...
Abstract Motivation The goal of pharmacogenomics is to predict drug response in patients using their single- or multi-omics data. A major challenge that clinical data (i.e. patients) with outcome very limited, creating a need for transfer learning bridge the gap between large pre-clinical datasets (e.g. cancer cell lines), as source domain, and target domain. Two discrepancies exist datasets: 1) input space, gene expression due difference basic biology, 2) output different measures response....