We shall simulate a clinical trial, analyse the data, and interpret the results together with a Jupyter notebook.
The source code of the fun project is available in a GitHub repo. And the essential information about the workshop is included in this one-pager PDF file (German), which can be shared and printed anywhere.
I want to thank Ralf Horstmoeller, Marie Pachtova, Paul Geser, David Gaul, Giulia Ferraina, and Jannick Lippuner for their support.
]]>Following last year’s success of the 1st Roche PMDA Summer School, we shall again open our door to PhD and outstanding master students enrolled in University Basel, ETH D-BSSE, and other Swiss Universities. To find out more about the summer school and to register, please visit our website at https://bedapub.github.io/PMDA-Summer-School.
All essential information is also included in this one-pager PDF file, which can be shared, printed, and posted anywhere.
Welcome to join us, and thank you for spreading the words!
]]>In this international event, the organizers expect around 70 participants. The event seeks to bridge bioinformatics and clinical research and have exciting talks, panel discussions and a poster session.
Registration is free of charge and participants are selected upon submitting an abstract. The submission site can be found here. Abstract submission deadline is July 17th, 2023.
Please find the flyer of the event. More information cab be obtained from the event’s website. Have fun with submission!
]]>The course series is organized by my colleague Adrian Roth, who is an expert of drug safety and new alternative methods. The course is attended mostly by master students of the Pharmacy Department of the University Basel. I have so far very good experience with the course: it covers a broad range of topics in drug discovery, it is given by a panel of experts working in diverse functions and assuming varying roles, the lecture is well organized by Adrian and her assistant Mrs Christiane Kocher, and probably the most important of all, the students are curious and engaged.
This semester, my talk will be about my current understanding of multiscale modeling of drug pharmacology and safety. My hope is to introduce three types of techniques that we often use to build computational models, namely mechanistic models, statistical models, and causal models. Together with in vitro, ex vivo, in vivo models, these modelling techniques give us the opportunity to understand how human body and drug interacts.
The key messages I tried to deliver are:
Here are the slides that I used for the lecture: slide deck. If you have comments, suggestions, and criticisms, please kindly let me know!
]]>This position is co-supervised by scientists in the newly established Roche Institute of Tissue Bioengineering (ITB) and the Predictive Modeling and Data Analysis (PMDA) Chapter in Pharmaceutical Sciences within Pharma Research and Early Development. The postdoctoral fellow will join our team at the interface of basic academic and translational research in the area of intestinal human model systems (HMS) and their application in drug metabolism and pharmacokinetics (DMPK) modeling. The postdoc will play a major role in the development of the intestinal HMS to reduce the current translational gaps with the current in vitro systems.
The postdoc fellow will:
Your profile
The job starts upon availability. A CV, a motivation letter, a publication list are required in your application. If you are interested, apply for the position here.
]]>In this position, the postdoc will process spatial transcriptomics (ST) datasets from mouse and human tissues with defined morphological structures and high relevance for toxicology readouts. The candidate will develop integrative analysis approaches of morphological patterns, cell identity and gene expression using deep learning methods. The output will deliver precise organ structure annotations and further provide cell type composition, spatial cell clustering, pathway activities and cell communication readouts. The postdoc will further help generate a ST database that provides an invaluable reference to assess compound-associated changes and to evaluate their translational relevance for patients.
This position is sponsored by the Roche Postdoctoral Fellowship Programme. The position is funded for two years, with the possibility of a third-year extension. The candidate works in a cross-functional spatial transcriptomics team consisting of members of the Predictive Modeling and Data Analytics (PMDA) and Pathology chapters, part of Roche Pharma Research and Early Development (pRED) Pharmaceutical Sciences department. The University of Heidelberg, Germany and the Swiss Institute of Bioinformatics, Lausanne, Switzerland are Roche’s academic partners for this Postdoctoral Fellowship.
In the position, you will
Requirements for the position:
The job starts in October 2022, or upon availability. A CV, a motivation letter, a publication list, and references are required in your application. If you are interested, apply for the position here.
]]>Population pharmacokinetic models are used in drug development to describe the time course of drug exposure in patients, and to investigate sources of variability in patient exposure. The technique can be used to simulate and assess alternative dose regimens.
Current strategy of population pharmacokinetic models is probably best described by a complex and iterative process that is manually performed by modelers. Automatic modelling instead uses algorithms to search for potential models that best describe a given dataset. Software tools are used to search the space of models and/or model features, to create and fit a series of candidate models, and to rank the models by pre-defined selection criteria.
In the position, you will
Requirements for the position:
If you are interested, apply for the internship here.
]]>We will close applications by June 17th and inform selected candidates at the beginning of July.
The topic of Roche PMDA Summer School 2022 will be single-cell gene expression in drug discovery. See the application page for more details.
As a pilot, the inaugural summer school is exclusively available to Ph.D. students associated with either University of Basel or ETH D-BSSE. In exceptional cases, applications of outstanding master students may be considered as well. Depending on the feedback, we may open the summer school to students of other academic institutes in the future.
If you meet the enrollment criteria, want to experience predictive modeling and data analytics in drug discovery, and have time and interest joining us between August 8th and 12th, 2022, please apply here: https://go.roche.com/PMDASummerSchool.
]]>Graph neural networks and causal inferences represent two state-of-the-art approaches to connectionist and symbolic learning. Graph neural networks, based on knowledge encoded in graphs and neural networks, are powerful tools for knowledge injection in machine learning and for representation learning (see the previous post on using knowledge-graph for patient-level data). Causal inference, based on Directed Acyclic Graphs (DAGs) and statistical models, is a powerful to discern causal effects from correlations caused by confounding (see a tutorial here on GitHub).
Our research question is whether we can find a midway that combines the advantages of both approaches - powerful prediction on the one hand, interpretability and counterfactual queries on the other hand - to predict and to understand mechanism and safety profiles of drugs. The premise is that drugs exert their efficacy and invoke adverse effects by interacting with and modulating targets and off-targets, which are components of biological networks. By combining our limited knowledge of biological networks and observational data, we wish to gain a better understanding of how and why drugs induce certain phenotypes by leveraging both computational approaches.
We welcome motivated bachelor or master students to apply. The internship starts latest in July and lasts 9 months. Please apply here on careers.roche.com or here on LinkedIn.
]]>The intern shall implement a bioinformatics workflow to assess the immunological relevance of our in vitro models. The workflow includes data management, data processing, and data analysis and visualization components. Once available, the workflow will allow us make better predictions about drug safety profiles and decisions about how to bring molecules to patients quickly and safely.
We expect you to apply for the internship if you meet all the requirements below.
Please apply for the position here.
You need to submit a CV, a motivation letter that does not include any personal details that refer to gender, age or ethnicity, and if available your open-source software repositories (for instance GitHub or GitLab) which we can access. Please don’t include an application photo.
The preferred start date of the 9-months-internship can be adjusted upon availability (ideally start before April 2022). Please clearly indicate your preferred start date and duration of the internship on your motivation letter.
]]>