Postdoc (m / f / d) in “Artificial intelligence and software development for behavioural quanti[...]

At the Max Planck Institute of Psychiatry, basic scientists and clinicians are working in close cooperation to understand the causes of complex diseases of the brain, searching for new prevention and treatment strategies. This translational research strategy is supported by a modern psychiatric research hospital as well as a number of research departments and research groups with expertise ranging from neurobiological research methods as well as clinical and behavioural studies to molecular genetics and computational approaches. The overarching goal is to optimise the benefit of research advances for the patients by way of close and innovative networking of basic research-oriented and clinical neuroscience.

The clinic is located in Schwabing between Luitpoldpark and the English Garden. It consists of three open and one sheltered psychiatric ward as well as the 60+ ward with a total of 120 beds for 2,000 patients per year and three day clinics. The Research Group “Statistical Genetics” led by Bertram Müller-Myhsok (https : / / www.psych.mpg.de / person / 33612 / 1495975) is looking as from 01.04.2024 for a motivated

Postdoc (m / f / d) in “Artificial intelligence and software development for behavioural quantification” Code : 8731

The successful applicant will work in close collaboration with the Research Group “ Stress Resilience ” led by Dr. Mathias Schmidt

Your tasks Key to the translational efforts described above is the accurate and detailed quantification of behavioural traits, which, after a long history of manual quantification, are currently being revolutionised by the incorporation of cutting-edge artificial intelligence.

The incorporation of deep learning algorithms into marker less motion tracking, with software such as DeepLabCut (https : / / www.mackenziemathislab.org / ) and SLEAP (https : / / sleap.ai / tutorials / tutorial.html), has paved the way for new ways of behavioural exploration. Both by means of training classifiers that can detect predefined traits in a supervised manner, and by the unsupervised embedding of tracked motion, scientists can now more reliably and efficiently evaluate what animals are doing during an experiment. It is in this setting that we recently developed and published DeepOF (https : / / www.nature.com / articles / s41467-023-40040-3), a Python package that is at the forefront of post-hoc analysis of pose estimation data for rodents.

The successful applicant would take the lead maintaining and developing tasks of DeepOF, and have freedom to develop new algorithms for supervised and unsupervised embedding of animal motion. The project will also delve into extending and generalising the software to new animals, adding multimodal integration models (such as with heart rate signals, neural activity, etc.), and closely collaborating with several international research groups using it at the time.

• PhD in biology, computer science, or a related field
• Experience with functional and object-oriented programming in Python (package development skills are a perk)
• Experience with deep learning frameworks (TensorFlow / PyTorch) is desired
• Experience with time series modelling / processing is desired
• Strong interest to understand the biological mechanisms underlying stress-related disorders
• Ability to work independently
• Excellent skills in spoken and written English

What we offer

• The multi-disciplinary environment in the research group fosters close collaborations between basic scientists as well as clinicians.
• Close contact with the DeepOF original developers, which will still be in the loop.
• State-of-the-art facilities, and a collaborative environment across statistics / computer science and behavioural science experts.
• Access to proper hardware to develop and train state of the art models, both in-house and provided by the Max Planck Computing & Data Facility.
• Travel budget to visit national and international conferences or workshops.
• The Max Planck Society offers a variety of soft skill and career-oriented courses for post-docs open to the successful applicant.
• Flexible work hours and the opportunity for occasional home office.
• Full time position (100%) for 24 months with a possibility for prolongation.
• Social benefits are paid in compliance with the regulations of the Max Planck Society.
• Very good connection to the Munich transport association (job ticket).

Your application :

We look forward to your online application with the following documents :

• A Cover letter, along with a brief summary of your previous projects, including applied techniques
• Your complete CV and contact details
• At least two references

The position is limited for two years.

Women are especially encouraged to apply and handicapped applicants with equal qualifications will be given preferential treatment.

Research Group led by Dr. Mathias Schmidt

Artificial Intelligence • Munich, Bavaria