- Antimicrobial Peptides and Activities
- Chemical Synthesis and Analysis
- Computational Drug Discovery Methods
- vaccines and immunoinformatics approaches
- Machine Learning in Bioinformatics
- RNA and protein synthesis mechanisms
- Machine Learning in Materials Science
- Monoclonal and Polyclonal Antibodies Research
- Biochemical and Structural Characterization
- Protein Structure and Dynamics
- Malaria Research and Control
- Immunotherapy and Immune Responses
- Click Chemistry and Applications
- T-cell and B-cell Immunology
- Chemokine receptors and signaling
- Bioinformatics and Genomic Networks
- interferon and immune responses
- Mobile Learning in Education
- Cytokine Signaling Pathways and Interactions
- Augmented Reality Applications
- Lipid Membrane Structure and Behavior
- Supramolecular Self-Assembly in Materials
- Virtual Reality Applications and Impacts
- Invertebrate Immune Response Mechanisms
- Drug Transport and Resistance Mechanisms
ETH Zurich
2015-2025
Board of the Swiss Federal Institutes of Technology
2011-2019
École Polytechnique Fédérale de Lausanne
2011-2018
Charles Humbert 8
2016
Goethe University Frankfurt
2007-2010
We present a generative long short-term memory (LSTM) recurrent neural network (RNN) for combinatorial de novo peptide design. RNN models capture patterns in sequential data and generate new instances from the learned context. Amino acid sequences represent suitable input these machine-learning models. Generative trained on could therefore facilitate design of bespoke libraries. RNNs with LSTM units pattern recognition helical antimicrobial peptides used resulting model sequence generation....
Abstract De novo drug design aims to generate molecules from scratch that possess specific chemical and pharmacological properties. We present a computational approach utilizing interactome-based deep learning for ligand- structure-based generation of drug-like molecules. This method capitalizes on the unique strengths both graph neural networks language models, offering an alternative need application-specific reinforcement, transfer, or few-shot learning. It enables “zero-shot"...
We have implemented the lecular esign aboratory's nti icrobial eptides package ( ), a Python-based software for design, classification and visual representation of peptide data. modlAMP offers functions molecular descriptor calculation retrieval amino acid sequences from public or local sequence databases, provides instant access to precompiled datasets machine learning. The also contains methods analysis circular dichroism spectra.The Python is available under BSD license URL...
Malaria blood stage parasites export a large number of proteins into their host erythrocyte to change it from container predominantly hemoglobin optimized for the transport oxygen niche parasite propagation. To understand this process, is crucial know which are exported cell. This has been aided by PEXEL/HT sequence, five-residue motif found in many proteins, leading prediction exportome. However, several negative (PNEPs) indicate that exportome incomplete and remains unknown if how further...
Many apicomplexan parasites, including Plasmodium falciparum, harbor a so-called apicoplast, complex plastid of red algal origin which was gained by secondary endosymbiotic event. The exact molecular mechanisms directing the transport nuclear-encoded proteins to apicoplast P. falciparum are not well understood. Recently, in silico analyses revealed second copy homologous components endoplasmic reticulum (ER)-associated protein degradation (ERAD) system organisms with plastids, malaria...
Constructive (generative) machine learning enables the automated generation of novel chemical structures without need for explicit molecular design rules. This study presents experimental application such a deep model to membranolytic anticancer peptides (ACPs) de novo. A recurrent neural network with long short-term memory cells was trained on α-helical cationic amphipathic peptide sequences and then fine-tuned 26 known ACPs by transfer learning. optimized used generate unique amino acid...
Understanding the structure–activity relationships and mechanisms of action membranolytic anticancer peptides could help them advance to therapeutic success.
Abstract The cell adhesion protein and tumour suppressor E-cadherin exhibits important functions in the prevention of gastric cancer. As a class-I carcinogen, Helicobacter pylori ( H. ) has developed unique strategy to interfere with functions. In previous studies, we have demonstrated that secretes protease high temperature requirement A (HtrA) which cleaves off ectodomain (NTF) on epithelial cells. This opens cell-to-cell junctions, allowing bacterial transmigration across polarised...
Abstract Natural products offer unexplored molecular frameworks for the development of chemical leads and innovative drugs. However, structural complexity natural compared with synthetic drug-like molecules often limits scaffold hopping potential natural-product-inspired design. Here we introduce a holistic representation incorporating pharmacophore shape patterns, which facilitates from to isofunctional compounds. This computational approach captures simultaneously partial charge, atom...
Abstract Membranolytic anticancer peptides represent a potential strategy in the fight against cancer. However, our understanding of underlying structure-activity relationships and mechanisms driving their cell selectivity is still limited. We developed computational approach as step towards rational design potent selective peptides. This machine learning model distinguishes between with without activity. classifier was experimentally validated by synthesizing testing selection 12...
Malaria parasites actively remodel the infected red blood cell (irbc) by exporting proteins into host cytoplasm. The human parasite Plasmodium falciparum exports particularly large numbers of proteins, including that establish a vesicular network allowing trafficking onto surface irbcs are responsible for tissue sequestration. Like P. falciparum, rodent berghei ANKA sequesters via irbc interactions with receptor CD36. We have applied proteomic, genomic, and reverse-genetic approaches to...
Abstract We present a “deep” network architecture for chemical data analysis and classification together with prospective proof‐of‐concept application. The model features self‐organizing map (SOM) as the input layer of feedforward neural network. SOM converts molecular descriptors to two‐dimensional image further processing. implemented lateral neuron inhibition contrast enhancement. achieved improved accuracy predictive robustness compared classifiers lacking layer. By nonlinear...
Augmented Reality (AR) tools are increasingly finding their way into education settings. Although use is still not widespread in educational contexts, the research literature indicates potential and effectiveness. However, overall specifically for sector there numerous gaps. This study investigates how of head-mounted AR displays such as Microsoft HoloLens can change learning what needs to be considered from a didactic perspective. The researched sample consists 18 student teachers with...
ABSTRACT Fimbriae of the human uropathogen Proteus mirabilis are only characterized surface proteins that contribute to its virulence by mediating adhesion and invasion uroepithelia. PMI2122 (AipA) PMI2575 (TaaP) annotated in genome strain HI4320 as trimeric autotransporters with “adhesin-like” “agglutinating adhesin-like” properties, respectively. The C-terminal 62 amino acids (aa) AipA 76 aa TaaP homologous translocator domains YadA from Yersinia enterocolitica Hia Haemophilus influenzae ....
De novo design of drug-like compounds with a desired pharmacological activity profile has become feasible through innovative computer algorithms. Fragment-based and simulated chemical reactions allow for the rapid generation candidate as blueprints organic synthesis.We used combination complementary virtual-screening tools analysis de designed that were generated aim to inhibit inactive polo-like kinase 1 (Plk1), target development cancer therapeutics. A homology model state Plk1 was...
Interfering with interferon: A low-molecular-weight inhibitor has been discovered that blocks the interaction between interferon-α (IFN-α) and its receptor (see picture for a model of interfaces). The resulting lead compound significantly reduces IFN-α production in vitro. NMR SPR experiments confirm direct IFN-α.
Designed peptides that bind to major histocompatibility protein I (MHC-I) allomorphs bear the promise of representing epitopes stimulate a desired immune response. A rigorous bioinformatical exploration sequence patterns hidden in mouse MHC-I allomorph H-2Kb is presented. We exemplify and validate these motif findings by systematically dissecting epitope SIINFEKL analyzing resulting fragments for their binding potential thermal denaturation assay. The results demonstrate only exclusively...
Abstract The computer‐assisted design and optimization of peptides with selective cancer cell killing activity was achieved through merging the features anticancer peptides, cell‐penetrating tumor‐homing peptides. Machine‐learning classifiers identified candidate that possess predicted properties. Starting from a template amino acid sequence, peptide cytotoxicity against range lines systematically optimized while minimizing effects on primary human endothelial cells. computer‐generated...