PHUSE US – Connecting F2F once more & industry collaboration more fertile than ever

Myself, Chris Price, James Black, Tad Lewandoswki, Harper Forbes & Kiran Bonda represented Roche/Genentech at the recent PHUSE US Connect in Atlanta. We presented a wide range of topics between us covering the pharmaverse, NEST, admiral, learnings from open source collaborations, leadership in an agile world, and automation in analysis results development.

The conference covered a diverse range of topics that impact Data Scientists from soft skills (or better described in one of the keynotes as “hard skills”) like professional development and leadership to more technical topics such as data standards, submissions, real world evidence, risk based monitoring, automation, as well as coding (more on that later). An increase from previous years in the number of panel discussions and workshops meant that there were more opportunities to take part in interactive discussions with industry peers which was very informative.

A major reflection of the conference was the wealth of topics around R and open source in particular (the open source story was standing room only on Monday)! PHUSE has definitely evolved beyond the primarily SAS focused talks of the past, and it was exciting to see how many other companies are serious about the road ahead to multi-lingual programming. With almost every R topic presented came the inevitable question from the crowd “When will this open source?”, which really speaks to the movement that Roche and others are heavily influencing. The conference also showed a great spirit of collaboration and open sharing of experiences around R, as many companies currently face similar challenges in scaling up R as a production solution: ranging from infrastructure, validation, implementation, and change management.

Another highlight was the CDISC partner sessions where CDISC presented both their new CDISC Open Rules Engine (CORE) and the draft Analysis Results Metadata (ARM) standard. It was a chance to hear directly from CDISC and to provide direct feedback on the details of the new areas that they are seeking to address outside of the CDISC Interchanges. Both of these efforts will have an impact moving forward on our work in Data Sciences.