- Will Spooner on Genomics England (co-meeting)
NOTE: not at our usual venue! In this very special Bioinformatics.London meetup, we're joining with the London Clinical and Health Data Science Meetup. The speaker will be Will Spooner from Genomics England. See: https://www.meetup.com/LonClinDatSci/events/257703253/ PLEASE NOTE: This event is limited to 50 free tickets at the below link. If you do not obtain a ticket you will not be admitted. https://www.eventbrite.co.uk/e/will-spooner-genomics-england-tickets-54281111265 Dr Will Spooner - Genomics England - "The power of genomics to transform clinical data science; accelerating biomedical research and changing the way the health system works" ABSTRACT: The NHS Long Term Plan released this month contained an exciting commitment: "Digitally-enabled care will go mainstream across the NHS" resulting in "an NHS where access to secure linked clinical, genomic and other data will support new medical breakthroughs and consistent quality of care". Genomics England has delivered a prototypical resource that delivers just that; an environment that provides access to linked clinical and genomic data from the 100,000 Genomes Project. Our Research Environment has been available to accredited researchers from both academia and industry for the past year. This talk will introduce the Research Environment and our development roadmap, and speculate about its role as a blueprint for clinical data science platforms of the future.
- Bioinformatics: TBD
Keep the date! Talks to be announced! venue to be confirmed but we'll work out somewhere. Is ticketing is required, the event will still be free but use EventBrite. Approximate schedule: 6:30: doors open 6:45: chitchat, announcements 7:00: talk starts 7:45: talk ends, adjourn to pub
- Graphs and knowledge bases with Grakn
NOTE: slight change of venue!!!! We'll be in the glamourous boardroom and data observatory of the Data Science Institute at Imperial College. Grakn allows you to model your domain using the well-known Entity-Relationship model at its full expressivity. Unlike other modelling languages, Grakn allows you to define type hierarchies, hyper-entities, hyper-relations, and rules to build rich knowledge models. Grakn isn't just a graph but a knowledge base that can be queried and manipulated. Tonight Tomas will show us how it can be employed in biomedical domains for reasoning and associations. Google map: https://goo.gl/maps/9q5DofRwrtL2 We'll adjourn to the Queens Arms afterwards: https://goo.gl/maps/omAqzqkwAS22 Approximate schedule: 6:30: doors open 6:45: chitchat, announcements 7:00: talk starts 7:45: talk ends, adjourn to pub
- Open Data Science Conference (ODSC): Europe 2018
Hi Everyone! Thanks for being a part of our community. We are thrilled to announce a partnership with The Open Data Science Conference (ODSC). See details from the event sponsors below - including a special offer just for our community members! ---------------------------------------------------------- I'm going to "#ODSC Europe2018,Open #DataScience Conference (Enter code ODSC40 & Save 40% or EUR18AC & Save 60% for academics, EUR18ST for Startups 60%)". 40% Off Limited Online Ends August 17th See you there? https://www.eventbrite.com/e/odsc-europe2018open-data-science-conferenceenter-code-odsc40save-40-tickets-42563659050?aff=estw … and https://odsc.com/london Here is the link to the Job Fair free registration: https://www.eventbrite.com/e/odsc-europe-job-fair-2018-tickets-48558659257 The ODSC team have offered a 63% student discount on any ticket until Friday, August 14th with code student63 or click on this link: https://www.eventbrite.com/e/odsc-europe2018open-data-science-conferenceenter-code-odsc40save-40-tickets-42563659050?discount=student63 ---------------------------------------------------------- The Open Data Science Conference is Europe’s leading data science conference. Over 4 days get access to 48 total tracks, 84 workshop/training sessions, 200+ presentations and, +1,800 attendees. The ODSC Europe 2018 schedule is now live. Hosted in London on September 19th - 22nd, the event attendees have a unique opportunity to get hands-on training from some of the world’s renowned data science experts including: The ODSC team have offered a 60% discount on any ticket until Friday, August 14th with code EUR18NP or click on this link. ● Andreas Mueller, PhD | Introduction to Machine Learning ● Juliet Moreiro Bockhop | The Path to Deep Learning with TensorFlow + Keras ● Michael Schmidt, PhD | Feature Engineering for Time Series Data ● Jared Lander | Machine Learning in R. Part 1 ● Aniket Anand Deshmukh | Transfer and Multi-task Learning ● Yves Hilpisch, PhD | Algorithmic Trading with Machine Learning and Deep Learning ● Andreas Mueller, PhD | Advanced Machine Learning with sci-kit learn ● Jan Willem Tulp | Learn D3 Essentials to get Started with Data Visualization ● And 40+ more… Totaling over 180 hours of applied talks, workshops, and training session, we offer the widest choice of the most important topics in data science. ODSC Europe 2018 will also feature: ● 47 introductory to advanced sessions in machine learning, deep learning, predictive analytics, data visualization, quant finance, and more ● Our Accelerate AI Summit | A full day of AI innovation, discussions for business professionals, executives, and startups looking to disrupt industry with AI ● Full day job fair to help you find your next career move ● Special events like Meet the Experts, Meet the Speaker, Dinner with Data Scientists, Networking Receptions, and more… Looking forward to meeting you all in London. The ODSC Team.
- Experiencing Information in Life Science Software
Eventbrite ticket Please sign-up using eventbrite so that you can get into Hays Life Sciences Recruitment building https://www.eventbrite.com/e/experiencing-information-in-life-science-sofware-tickets-47531992469 ABSTRACT Data is at the core of all science. As good user experience (UX) gains more traction within the life sciences community and an increasing recognition from life sciences executives, it is important to focus on the treatment of data as a key consideration in user experience. We call this information experience. Information experience is the design of visualisations, workflows, and features that facilitate user understanding and interaction with data. This includes data flows, manipulation, analysis, and reporting. Consideration of information experience is crucial to building effective life science sofware. This process ensures that scientist users can understand and trust the working of sofware tools. Most importantly it means that opportunities for discovery are not missed, maximising the value in data. In this talk we will discuss good practice for information experience in life science sofware. This will be based around a series of case studies, highlighting different methods of providing a good information experience to scientific users. Topics discussed will include methods to help users understand and interact with algorithmic analyses of data, visualisation aimed at encouraging discovery and hypothesis making, and accessible data reporting, helping expert and non-expert users to understand clinical data. SPEAKER NAME Simon Hazelwood-Smith SPEAKER BIO In his career as a UX designer and researcher, Simon has helped genomic startups define information experiences for their data platforms, collaborated with hospital teams to improve information exchange for clinical genetics, designed funding programs for the European Commission and used community labs to test the usability of handheld genetic sequencers. Simon works as a Designer and Researcher at the BioDesign team at Science Practice in London. He has previously worked in both research and diagnostic genetics laboratories, and has degrees in Genetics and Science and Innovation Policy. Approximate schedule: 6:30: doors open 6:45: chitchat, announcements 7:00: talk starts 7:45: talk ends, adjourn to pub
- RAPIDS 2018 - Reproducibility and Provenance in Data Science
More information and Signup at Eventbrite: https://www.eventbrite.co.uk/e/rapids-2018-reproducibility-and-provenance-in-data-science-tickets-46185840091 Join us at RAPIDS, a free one-day conference about how you can achieve reproducibility and provenance in your data science workflows through use of techniques and tooling. Find out how controlling your model, data and environment can give you control over your work that leads to greater productivity, collaboration and sanity! Meet 150 data science professionals, hear talks from leading practitioners and get hands-on in our workshops that give you skills that you can immediately apply to your work.
- Kai Davenport on Dotmesh
Yes, we're tardy but there is a meeting and we will get a better description. Kai Davenport will be talking about DotMesh: https://dotmesh.io/ "Dotmesh captures snapshots from stateful Docker and Kubernetes applications as ’datadots‘, and gives you the ability to label, store and share them as you would with version-controlled code" Ticketing is required (via EventBrite just to get names for security purposes), but the event will still be free: https://www.eventbrite.com/e/kai-davenport-on-dotmesh-tickets-46143554614 Title: Using dotmesh for reproducibility and provenance in data science. When conducting research and processing data, software gives us an immense range of benificial tools. We can run these tools locally or use cloud providers to run code on many high-powered machines. However, tools that manage the data we are working with are somewhat lacking. How many times have you made a copy of your dataset with a slightly different name or been unable to share reproducible experiments because of a lack of data management tools? In this talk, Kai will outline these problems and introduce dotmesh, a tool that enables the management of potentially large datasets with a particular focus on the upcoming release of "dotscience" - a suite of tools focusing on data-science. Being able to share the environment, code **and** data used to conduct an experiment is vital for the reproducibility and provenance of that experiment. Using modern tools such as git for code and AWS for infrastructure only get us part of the way there. dotmesh is an open source tool that we’ve developed which we think starts to address this third dimension of "data management". Speaker: Kai Davenport (@kai_davenport) Kai is an engineer for dotmesh and likes to stay on his toes doing work in different aspects of the stack, including ops (with k8s & gitlab), backend services (node.js, golang) and frontend development (JS, React, Redux). In a previous role, he was a developer advocate for ClusterHQ and continues his involvement in the world of storage by his work for dotmesh and writing the occasional blog-post for Portworx. He also run's an Ethereum meetup in Bristol. Approximate schedule: 6:30: doors open 6:45: chitchat, announcements 7:00: talk starts 7:45: talk ends, adjourn to pub
- Ontologies mapping tools and services
Ian Harrow for the Pistoia Alliance Ontologies Mapping Project on: Ontologies can include hierarchical relationships; taxonomies; Biomedical ontologies are critical for effective scientific search and data integration, and much effort has been invested in developing a broad range of valuable public ontologies. These cover many domains, and even within one domain there are many distinct but overlapping ontologies (e.g. disease & phenotype), the power of which can be increased if they can be combined. The Pistoia Alliance Ontologies Mapping (OM) project was created to support or develop better tools or services for mapping between ontologies in the same domain and to establish best practices for ontology management in the Life Sciences. In this talk Ian Harrow will present on 1) guidelines for selection of ontologies prior to application, 2) evaluation of top performing ontology mapping tools and 3) development of a prototype ontologies mapping service. Ticketing is required, the event will still be free but use EventBrite: https://www.eventbrite.com/e/ontology-mapping-with-pistoia-tickets-42229437384 Approximate schedule: 6:30: doors open 6:45: chitchat, announcements 7:00: talk starts 7:45: talk ends, adjourn to pub
- Managing & analysing data with Databiology
Databiology orchestrates biomedical analysis across with any data, any application, and on any infrastructure. Luke Smith will explain how we designed it to work with distributed data, a constantly changing landscape of bioinformatics tools, whilst ensuring scalability and auditability. http://www.databiology.com/ Databiology is a biomedical information management and process orchestration platform for the life sciences and healthcare industries that provides flexibility, freedom and peace of mind for research and development teams. Researchers use Databiology to be able to search across many different data sources with different owners. Application Publishers use Databiology to build agile analysis pipelines easily without the need to conform to proprietary API or workflow manager constraints. System administrators use Databiology to manage access to applications, and to multiple different compute and storage infrastructure side-by-side in order to provide users with a seamless experience. Databiology enables teams to configure, command and collaborate on projects that produce or consume biomedical data. Ticketing is required, the event will still be free but use EventBrite. because we need everyone's names for security purposes: http://www.eventbrite.com/e/managing-analysing-data-with-databiology-tickets-42229410303 NOTE: we're in Hays but in a different room to the usual. Look for Ben or his little helper when you enter Hays, and they'll direct you to the right place. Approximate schedule: 6:30: doors open 6:45: chitchat, announcements 7:00: talk starts 7:45: talk ends, adjourn to pub
- Graph Genome Tools and Snakemake
Ticketing is required (via EventBrite just to get names for security purposes), but the event will still be free. https://www.eventbrite.com/e/graph-genome-tools-and-snakemake-tickets-40168318522 There will be two talks: Morten Kallberg (Seven Bridges) - "Graph Genome Tools for Precision Medicine" Seven Bridges’ Graph Genome technology represents a fundamental rethinking of data structures to assess genomic variation at the population level. To address the inherent limitations of linear reference genomes, we have reconstructed the reference genome in a graph structure where each variant is represented as a branch in a directed acyclic graph. Unlike standard linear references, this structure makes use of information from an entire population to characterize genetic variants with unprecedented accuracy. Our tools learn from every new person sequenced, meaning that the graph-based reference improves with each additional genome. Using whole genome data from the PrecisionFDA truth challenge for benchmarking, we were able to identify variants with greater accuracy than all competing methods. Further, our production pipeline is faster and more cost-effective than competing whole-genome pipelines, thus facilitating population-scale analysis. and Paul Agapow (Imperial College) "Snakemake, for humane workflows & reproducibility" Have a complicated multi-step analysis? Tired of cryptic or jury-rigged pipelines? The Snakemake is a tool to create reproducible and scalable data analyses via a human readable, Python based language. I demonstrate how I used it for a hideously complicated analysis to generate easily digestible results via a single commandline script. Some advice is given about where Snakemake is (and isn't useful) and some tips and tricks. Approximate schedule: 6:30: doors open 6:45: chitchat, announcements 7:00: talk starts 7:45: talk ends, adjourn to pub