Open Positions

PhD position on Abstraction and reasoning challenge: Create an AI capable of solving reasoning tasks it has never seen before

Universitat Pompeu Fabra and Prof. Gianni De Fabritiis, ICREA research professor, are looking to recruit a PhD student for a PhD on AI.

Description:

Can a computer learn complex, abstract tasks from just a few examples? Current machine learning techniques are data-hungry and brittle—they can only make sense of patterns they’ve seen before. Using current methods like reinforcement learning, an algorithm can gain new skills by exposure to large amounts of data, but cognitive abilities that could broadly generalize to many tasks remain elusive. This makes it very challenging to create systems that can handle the variability and unpredictability of the real world, such as domestic robots or self-driving cars. However, alternative approaches, like inductive programming, offer the potential for more human-like abstraction and reasoning.

The abstraction and reasoning corpus (ARC) provides a benchmark to measure AI skill-acquisition on unknown tasks, with the constraint that only a handful of demonstrations are shown to learn a complex task (example). This competition was initially created by the creator of the Keras neural networks library and it’s explained in this paper. The idea is to move beyond the competition timeframe to create an AI that can solve reasoning tasks it has never seen before and set up a path toward a PhD in AI. It is expected that novel work in terms of a paper should be produced during this period.

Location:

Barcelona, Spain, Barcelona Biomedical Research park. The laboratory is part of the Barcelona Biomedical Research Park which, with a privileged location on the shoreline of the Mediterranean sea, constitutes one of the most exciting interdisciplinary research centres in Southern Europe with more than 1000 scientists in the building alone.

Information on the application process:

Please send an email to gianni.defabritiis[at]upf.edu with subject “JOB PhD1 2020” with a CV and a cover letter together with the names of up to three contacts for requesting recommendations

Deadline to submit applications 15/6/2020

PhD and Postdoctoral Position on artificial intelligence at the Computational Science Laboratory, University Pompeu Fabra, Barcelona

Description:

Universitat Pompeu Fabra and Prof. Gianni De Fabritiis, ICREA research professor, are looking to recruit either a PhD student and a postdoc on the topic of applying machine learning and reinforced learning.

Relevant References:

Google Scholar

Aim:

This project aims to develop methods and applications that are transformative in the field of artificial intelligence by focusing on solving AI agent-based challenges (e.g. Unity 3D environments, but not only). The results are expected to be publishable in the highest impact scientific journals. There will be ample opportunities to present his/her work at international meetings and conferences. One example of such activity was the Unity obstacle challenge, where a player had to navigate a 3D environment solving multiple intelligence related tasks like navigation, using keys, boxes, etc. In such instance, the lab reaches second place: See this blog and also this one.

The computational science laboratory, led by Gianni De Fabritiis has been successfully applying deep learning approaches on fundamental problems in these areas, such as protein-ligand docking, affinity prediction and protein folding, winning several international challenges. Currently, we are focused on developing intelligent agents that can outperform humans on specific problems.

Project and Institution that finance the contract:

The work is supported by grant: AEI-BIO2017-82628-P (FEDER/UE) DE FABRITIIS from the Spanish Ministry of Economy (MINECO), FEDER and Agencia Estatal de Investigación sources and Acellera.

Official number reference:

AEI-BIO2017-82628-P (Aplicación de métodos de aprendizaje automático y de aumento de datos en el estudio de la conformación proteica y el reconocimiento molecular para el desarrollo de fármacos)

Skills and Experience:

The candidate will preferably have a profile in computer science, physics or mathematics. However, we seek exceptional candidates with a passion for computing, the ambition to work in very innovative projects more than specific profiles.

The capability to think out of the box, the ambition to work in very innovative projects and very good communication skills in English.

Previous experience in reinforcement learning and related fields, Python proficiency and coding skills, knowledge of Tensorflow or pytorch, familiarity with Linux and the ability to work with version control systems (e.g. git) are required.

Benefits of the opening:

Location

The laboratory is located in the Barcelona Biomedical Research Park which, with a privileged location on the shoreline of the Mediterranean sea, constitutes one of the most exciting interdisciplinary research centres in Southern Europe with more than 1000 scientists in the building alone.

Facilities

Access to state of the art computational resources and large amount of simulation data, which will be crucial for the development and validation of novel computational protocols. The lab is equipped with a cluster with 60 GPUs and has exclusive access to GPUGRID.net, a distributed computing project with 5000 GPUs.

Salary

We offer a competitive phd or postdoc salary, depending on experience.

Duration

3-4 years

Information on the application process

Please send an email to gianni.defabritiis[at]upf.edu with subject “JOB phd2” with a CV and a cover letter together with the names of up to three contacts for requesting recommendations

Deadline to submit applications:

31/7/2020

Closed Positions

PhD Position in machine learning for structural molecular biology

Academic supervisor:

Universitat Pompeu Fabra and Prof. Gianni De Fabritiis, Icrea research professor, are looking to recruit a young scientist for a PhD of the duration of approximatively 3 years on a interdisciplinar topic machine learning, machine intelligence, simulations and computational biology.

Aims:

This project aims to develop large scale artificial neural networks for machine intelligence applied to structural and computational biology. The aim is to go substantially beyond the state-of-the-art, exploring local update learning rules, unsupervised and reinforcement learning approaches.

We expect the candidate to participate in the development of new neuro computation approaches inspired by brain cortical circuits, active/passive dendrides and deep learning, in relation with GPU and distributed computing. By working in this project, the researcher will have access to state of the art computational project like GPUGRID.net and large amount of molecular simulation data, which will be crucial for the development and validation of novel computational protocols. This project is expected to lead to discoveries that will be publishable in the highest impact scientific journals.

Location:

Barcelona, Spain,Barcelona Biomedical Research park The laboratory is part of the Barcelona Biomedical Research Park which, with a privileged location on the shoreline of the Mediterranean sea, constitutes one of the most exciting interdisciplinary research centres in Southern Europe with more than 1000 scientists in the building alone.

Candidate Profile:

The candidate will preferably have a profile in computer science, mathematics, physics, and engineering, but other fields are also considered. We seek exceptional candidates with a passion for programming and computing, the capability to think out of the box, the ambition to work in very innovative projects and very good communication skills in English. Prior knowledge in neural information processing, deep learning is desirable. Most of the code will be likely written in C/C++ and Python.

Facilities:

The candidate will be able to use GPUGRID.net, one of the largest volunteer computing project worldwide with thousands of GPUs. The laboratory is very well equipped with access to a local GPU and CPU cluster, hundreds of TB of storage and experimental facilities.

The molecular modelling will be using state of the art software environments in molecular modelling and simulations (HTMD) developed by the research group in collaboration with the spin-off company Acellera.

There will be ample opportunities to present his/her work at international meetings and conferences and to collaborate with other research groups in USA and Europe.

Funding and Eligibility:

Interested candidates should apply by sending a complete CV and a short letter of presentation indicating their interests together with the names and addresses of three referees to: Gianni De Fabritiis (gianni.defabritiis at upf.edu). The title of this offer must be clearly stated in any communication.

Deadline: To be filled within 2016

Some Relevant References:

PhD Position in protein protein molecular recognition and design

Academic supervisor:

Universitat Pompeu Fabra and Prof. Gianni De Fabritiis, Icrea research professor, are looking to recruit a young scientist for a PhD of the duration of approximatively 3 years in Barcelona on the topic of protein-protein recognition using molecular simulations.

Aims:

This project aims to develop methods for characterizing molecular recognition between protein-protein and protein-antibodies. The aim is to go substantially beyond the state-of-the-art, exploring new methodologies based on advanced software platforms (www.htmd.org) with the aim to design novel potential therapeutic agents.

We expect the candidate to participate in the development and application of the methods in collaboration with clinical and experimental groups. By working in this project, the researcher will have access to state of the art computational resources using GPUGRID.net and large amount of molecular simulation data. This project is expected to lead to discoveries that will be publishable in the highest impact scientific journals.

Location:

Barcelona, Spain,Barcelona Biomedical Research park. The laboratory is part of the Barcelona Biomedical Research Park which, with a privileged location on the shoreline of the Mediterranean sea, constitutes one of the most exciting interdisciplinary research centres in Southern Europe with more than 1000 scientists in the building alone.

Candidate Profile:

The candidate will preferably have a profile in chemistry, biology, physics or related fields. We seek candidates with a passion for biology and computing, the capability to think out of the box, the ambition to work in very innovative projects and very good communication skills in English. Prior knowledge in molecular simulations, protein-design methods, knowledge-based methods is desirable, but we are really looking for self-motivated, exceptional candidates. Most of the methods will be developped using Python and HTMD. A good knowledge of programming in Python or other programming languages is therefore required.

Facilities:

The candidate will be able to use GPUGRID, one of the largest volunteer computing project worldwide with thousands of GPUs. The laboratory is very well equipped with access to a local GPU and CPU cluster, hundreds of TB of storage and experimental facilities.

The molecular modelling will be using state of the art software environments in molecular modelling and simulations developed by the research group in collaboration with the spin-off company Acellera.

There will be ample opportunities to present his/her work at international meetings and conferences and to collaborate with other research groups in USA and Europe.

Funding and Eligibility:

Interested candidates should apply by sending a complete CV and a short letter of presentation indicating their interests together with the names and addresses of three referees to: Gianni De Fabritiis (gianni.defabritiis at upf.edu). The title of this offer must be clearly stated in any communication.

Deadline: To be filled within 2016

Some Relevant References:

PhD Position in rational drug design of intrinsically disordered proteins

Academic supervisor:

Prof. Gianni De Fabritiis

Aims:

Coordination of cellular processes requires time-ordered formation of functional complexes for which the proper timing of binding and unbinding is critical. Although evidence exists for the relevance of kinetics for biological function, the underlying physical-chemistry mechanisms are still largely unexplored. This is especially true for protein domains without a native folded conformation as, in absence of a structured state, a shift of the equilibrium towards active or inactive states may not be feasible.

Using massively distributed molecular dynamics simulations, high performance codes and new Markov state analytical tools, all largely developed by us, we have shown in a series of pilot studies the capability to resolve molecular kinetics once the multi-millisecond timescale is reached. Preliminary mathematical kinetic models also show that, by modulating the molecular kinetics of competing on-pathway metastable states, it is possible to control the equilibrium of the complex, a novel and potentially important mechanism for intrinsically disordered proteins.

In this project, we aim to investigate the relevance and generality of molecular kinetics for intrinsically disordered protein domains and its modulation by small molecules. We will employ a computationally driven approach to resolve kinetics at atomic resolution, and will integrate our computational predictions with standard assays and experiments from collaborations with experimental experts and local facilities.

We expect the candidate to participate in the design and implementation of computational protocols to using the advanced computational developed in the group.

Location:

Barcelona, Spain, Barcelona Biomedical Research Park. The GRIB/UPF is part of the new Barcelona Biomedical Research Park which, with a privileged location on the shoreline of the Mediterranean sea, constitutes one of the most exciting interdisciplinary research centres in Southern Europe in one of the fastest growing biomedical research regions in the continent with over a thousand scientists. Candidate Profile:

The candidate will preferably have a profile in biochemistry, chemistry, pharmaceutics, biotechnology, biology. We also have in our group students with a background in computer science, physics, mathematics. Very good communication skills in English are required.

Facilities:

The candidate will be able to use GPUGRID which gives the possibility to run high throughput molecular dynamics simulations of multi millisecond timescales routinely. The laboratory is very well equipped with also access to a local GPU cluster and experimental facilities to measure affinities and kinetics.

The molecular modelling will be using state of the art software environments in molecular modelling and simulations (HTMD) developed by the research group in collaboration with Acellera.

Funding and Eligibility:

The project in supported by the Spanish ministry of science. The application process consist in sending a CV and a proof of your bachelor degree to the formal application page made available. The call is open to EU nationals. There will be ample opportunities to present his/her work at international meetings and conferences. Eligible candidates should apply by sending a complete and up to date CV together with the names and addresses of two academic referees to: Gianni De Fabritiis (gianni.defabritiis at upf.edu). The title of this offer must be clearly stated in any communication.

Deadline:

27th June 2015 CLOSED

References:

Further references are available here