Postdoctoral Fellows - Deep mutagenesis and machine learning for human genetics

Do you want to help us improve human health and understand life on Earth? Make your mark by shaping the future to enable or deliver life-changing science to solve some of humanity’s greatest challenges. About the Role: We have an exciting opportunity for up to three Postdoctoral Fellows to join the Lehner group for a 3 year fixed term contract. You will be performing cutting edge experiments focused on understanding, predicting and engineering genetic variation at scale using deep mutational scanning. You will be responsible for: Developing independent and collaborative research projects that apply selection assays at scale for the massively parallel quantification of biophysical and molecular activities of protein and RNAs. Using these data sets to address fundamental questions in biophysics and clinical genetics. Writing up manuscripts, communicating results through talking, presenting posters at scientific meetings Collaborate with internal teams and external collaborators to ensure project progress and provide regular updates Taking an active part in your development, including attending training and being an active member of our Postdoctoral Fellow community About You: You will have a background in genomics, biophysics or a related discipline and a desire to perform quantitative experiments at scale to address fundamental questions in molecular biology. Skills in programming, machine learning and statistics are desirable but can also be developed in the role. This opportunity will allow you to work on independent research whilst receiving training and support in both experimental and computational skills from a variety of sources. You will have excellent communication skills and the ability to communicate effectively with stakeholders. For more information about being a Postdoctoral Fellow with us, please visit: https://youtu.be/1-vxm2nBmFA About Us: The Lehner group is newly established at the Wellcome Sanger Institute. The lab is focused on combining massively parallel genetic perturbation experiments with modelling and machine learning to understand, predict and engineer how changes in DNA sequence alter the activities, interactions and regulation of proteins and RNAs. The goal is both to generate reference atlases and predictive models of variant effects to guide clinical genetics and datasets of sufficient scale and diversity to directly tackle the fundamental sequence-to-activity encoding problems of molecular biology using artificial intelligence and modelling. The long-term goal is to make molecular biology and genetics predictive and programmable. Individual projects are focussed on fundamental questions in biophysics and genetics or on important protein and gene families or diseases. Example include understanding, predicting and engineering allostery, receptors and membrane proteins, protein binding affinity and specificity, small molecule and drug binding, cancer genes, splicing and post-transcriptional regulation, transcription factors, and intrinsically disordered proteins. The resulting data and models will support clinicians in diagnosing diseases and understanding disease mechanisms. It will also support scientists in designing and producing new proteins and small molecules for disease treatment and bioengineering. Essential Skills: PhD in genomics, biophysics, quantitative, synthetic or molecular biology or a relevant non-biological discipline Ability to analyse and interpret genomic data with strong quantitative skills Scientific independence and an ability to solve problems An open and collaborative approach to research Ability to effectively prioritise, multi-task and work independently Ability to work to tight timelines Motivation and ambition to make a personal contribution to the Wellcome Sanger Institute research Ability to develop and maintain effective working relationships with wide range of persons of differing levels, abilities and knowledge Commitment to personal development and growth of knowledge and skills Thorough record keeping, with strong organisational and communication skills Demonstrates inclusivity and respect for all Excellent written and oral communication skills Other Information: Salary per annum: £35,707-£44,712 Application Process: Please apply with your CV and a Cover Letter outlining how you meet the criteria set out above. Please also indicate in your cover letter why you are interested in joining the Lehner group. Closing Date: 16th February 2023 Working at Wellcome Sanger: Our flexible-hybrid working environment is designed to support a healthy work-life balance. This means you can work flexibly with a combination of working from home, and working from our Campus to allow you to focus on being productive and part of the team while enjoying the benefits of working flexibly. We aim to attract, recruit, retain and develop talent from the widest possible talent pool, thereby gaining insight and access to different markets to generate a greater impact on the world. We have a supportive culture with the following staff networks, LGBTQ+, Parents and Carers and Race Equity to bring people together to share experiences, offer specific support and development opportunities and raise awareness. The networks are also a place for allies to provide support to others. We want our people to be whoever they want to be because we believe people who bring their best selves to work, do their best work. That’s why we’re committed to creating a truly inclusive culture at Sanger Institute. We will consider all individuals without discrimination and are committed to creating an inclusive environment for all employees, where everyone can thrive. Our Benefits: We are proud to deliver an awarding campus-wide employee wellbeing strategy and programme. The importance of good health and adopting a healthier lifestyle and the commitment to reduce work-related stress is strongly acknowledged and recognised at Sanger Institute. Life at the Sanger Institute is unique. We are tackling some of the most difficult challenges in genomic research. Our people are shaping the future by delivering life-changing science with the reach, scale, and creativity to solve some of humanity’s greatest challenges. We aim to attract, recruit, retain and develop talent from the widest possible talent pool, thereby gaining insight and access to different markets to generate a greater impact on the world.

Google Colaboratory

Itai Yanai on Twitter

“The result from the data analysis only has a chance of being right if one has explored every possible way that it might be wrong.”

ClimaX: A foundation model for weather and climate

Most state-of-the-art approaches for weather and climate modeling are based on physics-informed numerical models of the atmosphere. These approaches aim to model the non-linear dynamics and complex interactions between multiple variables, which are challenging to approximate. Additionally, many such numerical models are computationally intensive, especially when modeling the atmospheric phenomenon at a fine-grained spatial and temporal resolution. Recent data-driven approaches based on machine learning instead aim to directly solve a downstream forecasting or projection task by learning a data-driven functional mapping using deep neural networks. However, these networks are trained using curated and homogeneous climate datasets for specific spatiotemporal tasks, and thus lack the generality of numerical models. We develop and demonstrate ClimaX, a flexible and generalizable deep learning model for weather and climate science that can be trained using heterogeneous datasets spanning different variables, spatio-temporal coverage, and physical groundings. ClimaX extends the Transformer architecture with novel encoding and aggregation blocks that allow effective use of available compute while maintaining general utility. ClimaX is pre-trained with a self-supervised learning objective on climate datasets derived from CMIP6. The pre-trained ClimaX can then be fine-tuned to address a breadth of climate and weather tasks, including those that involve atmospheric variables and spatio-temporal scales unseen during pretraining. Compared to existing data-driven baselines, we show that this generality in ClimaX results in superior performance on benchmarks for weather forecasting and climate projections, even when pretrained at lower resolutions and compute budgets. The source code is available at https://github.com/microsoft/ClimaX.

Percy Liang on Twitter

“I worry about language models being trained on test sets. Recently, we emailed [email protected] to opt out of having our (test) data be used to improve models. This isn't enough though: others running evals could still inadvertently contribute those test sets to training.”

Krishna Shenoy, engineer who reimagined how the brain makes the body move, dies at 54

Six Faculty Positions at UBC in Biological Resilience  | Life Sciences Institute

Multi Career-stage, Diversity Cluster Hire  The Life Sciences Institute’s (LSI) recent designation as a Global Research Excellence Institute at the University of British Columbia (UBC), prompted the launch of the Biological Resilience Initiative.