Postdoctoral Appointee - Data Science and Learning for End-to-End Biological Programming
·
Argonne National Laboratory
·
Lemont, IL
SessionJob Fair
Event Type
Job Posting
TP
W
TUT
XO
TimeMonday, 9 November 20209am - 8pm EDT
Location
DescriptionPosition Description:
The Data Science and Learning (DSL), Biosciences (BIO), and X-ray Science (XSD) divisions at Argonne National Laboratory invites applications for three postdoctoral researchers for developing scalable statistical inference approaches for programming biological systems. Biological programming applications (e.g., protein design, synthetic biology) are now increasingly automated through robotic instrumentation and cloud-lab platforms (e.g., Emerald Lab). Our group is particularly interested in exploiting these high throughput capabilities in expanding how biological programming is implemented in an end-to-end fashion – i.e., multiple iterations of design, build, test and learn all run autonomously through AI-enabled infrastructure.
While many biological programs seek to abstract the design-build-test-learn (DBTL) cycle, they need scalable artificial intelligence (AI) and machine learning (ML) methods for: (i) representational learning and symbolic reasoning to infer biological abstractions from existing data; (ii) active learning for incorporating current (or partial) observables into biological abstractions to yield better design of experiments; and (iii) probabilistic model checking to enable robust execution of biological programs. We propose to capture these activities as platform for designing robust end-to-end biological programs for synthetic biology applications.
We are interested in bringing together an interdisciplinary team that has expertise in AI, high performance computing, robotics and systems/structural biology to explore the combinatorics of design involved in studying self-assembly processes with intrinsically disordered proteins and targeting secure biosystems design.
The candidates will build scalable AI/ML models on data center AI systems (e.g., Cerebras CS-1 ML accelerator and Argonne's Aurora exascale supercomputer) that are tightly integrated with mechanistic modeling frameworks (e.g., MD simulation platforms as well as flux-balance analysis tools), and implement end-to-end experimental design on Argonne’s robotic platforms.
The opportunity will involve close collaborations with researchers at the University of Chicago and Northwestern University along with other national laboratories including Brookhaven, Livermore, and Berkeley.
Position Requirements:
PhD in a computer science, physical sciences or engineering or related field.
Comprehensive experience programming in one or more programming languages, such as C, C++, and Python
Experience with machine learning methods and deep learning frameworks, including tensorflow, pytorch.
Experience in heterogenous programming and GPU programming in the machine learning context is valuable.
Software development practices and techniques for computational and data-intensive science problems.
Exceptional communication skills, ability to communicate effectively with internal and external collaborators and ability to work in team environment.
Ability to model Argonne’s Core Values: Impact, Safety, Respect, Integrity, and Teamwork.
Preferred:
Experience and understanding of how synthetic biological circuits function (basic understanding of CRISPR/Cas9 mechanisms)
Experience in working with structural biology applications (especially with disordered proteins).
Experience in applied machine learning (e.g., successful projects that used ML to solve scientific problems).
Experience with high-performance computing and handling robotic platforms (e.g., Hudson Robotics).
Ability to provide project leadership.
The Data Science and Learning (DSL), Biosciences (BIO), and X-ray Science (XSD) divisions at Argonne National Laboratory invites applications for three postdoctoral researchers for developing scalable statistical inference approaches for programming biological systems. Biological programming applications (e.g., protein design, synthetic biology) are now increasingly automated through robotic instrumentation and cloud-lab platforms (e.g., Emerald Lab). Our group is particularly interested in exploiting these high throughput capabilities in expanding how biological programming is implemented in an end-to-end fashion – i.e., multiple iterations of design, build, test and learn all run autonomously through AI-enabled infrastructure.
While many biological programs seek to abstract the design-build-test-learn (DBTL) cycle, they need scalable artificial intelligence (AI) and machine learning (ML) methods for: (i) representational learning and symbolic reasoning to infer biological abstractions from existing data; (ii) active learning for incorporating current (or partial) observables into biological abstractions to yield better design of experiments; and (iii) probabilistic model checking to enable robust execution of biological programs. We propose to capture these activities as platform for designing robust end-to-end biological programs for synthetic biology applications.
We are interested in bringing together an interdisciplinary team that has expertise in AI, high performance computing, robotics and systems/structural biology to explore the combinatorics of design involved in studying self-assembly processes with intrinsically disordered proteins and targeting secure biosystems design.
The candidates will build scalable AI/ML models on data center AI systems (e.g., Cerebras CS-1 ML accelerator and Argonne's Aurora exascale supercomputer) that are tightly integrated with mechanistic modeling frameworks (e.g., MD simulation platforms as well as flux-balance analysis tools), and implement end-to-end experimental design on Argonne’s robotic platforms.
The opportunity will involve close collaborations with researchers at the University of Chicago and Northwestern University along with other national laboratories including Brookhaven, Livermore, and Berkeley.
Position Requirements:
PhD in a computer science, physical sciences or engineering or related field.
Comprehensive experience programming in one or more programming languages, such as C, C++, and Python
Experience with machine learning methods and deep learning frameworks, including tensorflow, pytorch.
Experience in heterogenous programming and GPU programming in the machine learning context is valuable.
Software development practices and techniques for computational and data-intensive science problems.
Exceptional communication skills, ability to communicate effectively with internal and external collaborators and ability to work in team environment.
Ability to model Argonne’s Core Values: Impact, Safety, Respect, Integrity, and Teamwork.
Preferred:
Experience and understanding of how synthetic biological circuits function (basic understanding of CRISPR/Cas9 mechanisms)
Experience in working with structural biology applications (especially with disordered proteins).
Experience in applied machine learning (e.g., successful projects that used ML to solve scientific problems).
Experience with high-performance computing and handling robotic platforms (e.g., Hudson Robotics).
Ability to provide project leadership.
·
·