Compute Canada

Coast to Coast Seminar Series

The Coast to Coast Seminar is an hour-long presentation given on a scientific topic and made accessible to audiences at a number of remote sites through collaboration technology. C2C seminars are held every two weeks throughout the academic year alternating between the West Coast and the East Coast of Canada.

The topic of the Fall 2014 C2C seminar series is "Deep Sequencing Antibody and T-cell Receptor Repertoires for the Study of Infectious and Autoimmune Disease, and Development of Vaccines and Therapeutics".

This Fall 2014 Coast to Coast Seminar Series will examine an immediate and pressing “Big Data” problem faced by researchers examining the immune response to infectious and autoimmune diseases, and those developing vaccines and therapeutic antibodies. The series is partly motivated by a specific middleware prototype that The IRMACS Centre is developing called iReceptor, supported by a CANARIE contract. More importantly, the proposed series will be an important opportunity to help integrate the Canadian and greater international community that is grappling with this Big Data problem.

Only a few years ago, a researcher attempting to characterize the human immune response would have been able to sequence only a few hundred Antibody/B-cell receptor or T-cell receptor sequences per patient. This sequencing might have been conducted as part of research on infectious (e.g., flu, HIV) or autoimmune diseases (e.g., Multiple Sclerosis, Type 1 Diabetes), or during the development of vaccines or therapeutic antibodies. However, with the application of “deep sequencing” or “next generation sequencing” (NGS) techniques, it is now possible to sequence millions of these sequences per individual, per time point (e.g., early or late in infection, before or after vaccination), and from different sets of cells within the body. This is a prime example of the challenge of “Big Data.” In order to optimize the utility of these data for biomedical research and patient care, it will be critical to store, share and compare these huge databases. At present there is no integrated system to easily conduct these comparisons. Such a system will have to present a common data format, for both the sequence data and the sample metadata such as ethnicity, genetic background, treatment regime, gene expression, and clinical outcome, to highlight just some of the supporting data. The ability to share and compare these data is critical to using them optimally for biomedical research and patient care, and will involve significant hurdles in terms of ethics and consent, confidentiality and data security, intellectual property. This is the purpose of the workshops that are being organized in Vancouver, and a Coast to Coast seminar series addressing this topic will help to integrate the Canadian and greater international community as one step in this initiative.

Please contact us at help@hpcvl.org if you want to attend one of these seminars.

The first lecture in the series will be broadcast on September 30, 2014

Details can be found at http://www.irmacs.sfu.ca/events/coast-coast-seminars.


September 30, 2014 2:30 pm - 3:30 pm

Coast to Coast Seminar Series - Tools for the Analysis of NGS Databases of Antibody Repertoires

Andrew Bradbury – Group Leader, Los Alamos National Laboratories, New Mexico

Abstract:

...Coming Soon...

About the Lecturer:

...Coming Soon...


October 14, 2014 2:30 pm - 3:30 pm

Coast to Coast Seminar Series - The Potential for NGS Databases of Antibody Repertoires for the Pharmaceutical Industry

Jacob Glanville - Chief Science Officer, Distributed Bio, Seattle

Abstract:

...Coming Soon...

About the Lecturer:

...Coming Soon...


October 28, 2014 2:30 pm - 3:30 pm

Coast to Coast Seminar Series - Deep Sequencing T-cell Receptor Databases For Developing Novel Treatments of Ovarian Cancer

Rob Holt - Director of Sequencing, BC Michael Smith Genome Sciences Centre, Vancouver

Abstract:

...Coming Soon...

About the Lecturer:

...Coming Soon...


November 25, 2014 2:30 pm - 3:30 pm

Coast to Coast Seminar Series - HIV/SLE Receptor Repertoire Comparisons: How Would Integrated Data Systems Help?

Jamie Scott - Tier I Canada Research Chair in Molecular Immunity, SFU

Abstract:

...Coming Soon...

About the Lecturer:

...Coming Soon...

ADF Density-Functional Software

HPCVL offers access to the ADF computational chemistry package. This includes the Amsterdam Density Functional (ADF) code for molecular DFT, as well as the BAND solid-state software, and other components. The software runs on our main M9000 comp[ute cluster. To use this software, please type or include in your setup files the command:

use adf2012.01

Alternatively, you can type

use adf

and will be set up for the most current version. If you need to continue to use earlier versions, you can do so by typing

use adf2008.01 or
use adf2009.01 

Details about using ADF on our systems can be found in the FAQ file.

Please let us know if you encounter any problems with the new version or need assistance running ADF jobs.

System Changes after Storage Upgrades

Apart from a substantial increase in available disk space, some other alteration have been made to our systems:

  • The new default for MPI programs is now ClusterTools 8.1 instead of CT 7.1. This means that using mpirun and compiler macros such as mpif90 or mpicc will use CT8 now. If you need to stick with CT7, you need to issue a "use ct7" command to override the default. The use of CT6 with mprun and mpif90 etc. is not affected by this.
  • We have added the queues for our Enterprise M9000 cluster to the default Grid Engine queue pool, so that jobs that were submitted without additional specification will be either scheduled on the Sunfire 25K's or the Enterprise M9000 servers.

    Note, that if a specific cluster is desired (for instance, if you have compiled code specifically for the M9000 servers) it can be requested by inserting a line:

    #$ -l qname=x.q

    into the Grid Engine script, where x.q is 25k.q for the Sunfires and m9k.q is for the M9000's.

External Usage Stats

HPCVL provides resources to many researchers external to the four CFI applicant institutions. The graph below shows external usage of the CFI funded CPU cluster at the central site. The Canada Foundation for Innovation (CFI),the Ontario Innovation Trust (OIT), and the Ontario Ministry of Research and Innovation have funded the resources at that site and the usage is represented as the percent of the usage of that equipment. Currently, 40 CPUs of the 1008 CPU cores Sun Fire cluster are set aside as workup and test resources and are not included in the percent usage of the cluster.

The total number of installed CPUs in the Sun Fire cluster is 1008 with the additional CPUs being funded through the Sun Microsystems. The total disk storage capacity is 160 TB of T3 StorEdge.

External CFI Usage

Comparing MPI and OpenMP

An interesting article comparing shared and distributed-memory approaches to parallel computing can be found here.

HPCVL strikes partnership with NAG

On November 16, 2002 HPCVL announced a partnership with the Numerical Algorithms Group. Please check here for the press release.