A web-based application for querying, browsing and downloading small molecule biological screening results

Technologies: OWL, Jena, Java, Javascript, MySQL
Website: baosearch.ccs.miami.edu

You can participate locally through The Lab Miami HACK FOR CHANGE: MIAMI – ARTS AND MORE.  The event takes place on Saturday, June 1, 2013 – 9:00am to Sunday, June 2, 2013 – 5:00pm, at 400 NW 26th Street, Miami  Florida  33127

Register by visiting  http://hackformiami.eventbrite.com/

Cancer is a multifactorial disease with a striking heterogeneity due to genetic, epigenetic and transcriptional changes involving a myriad of genes and proteins. While these factors are relevant to clinical prognosis and medical treatment of patients, a system’s approach is needed to unravel the complexities underlying intertwining carcinogenesis mechanisms. Given accurate experimental measurements, the presence at multiple scales of stochastic dynamics involved in gene regulation and protein‐protein interactions (PPI) requires that both the analysis of differential (cancer versus normal) conditions and the treatment of the associated uncertainty are taken at combined omics‐scale levels. In particular, networks allow for the straightforward integration between molecular, genetic, clinical and topological features in a unifying context. Models can then be built from network‐embedded measurable values to assess the variation significantly affecting the cellular mechanisms involved in cancer. By treating cancer as a systems disease, powerful computational instruments become available, and especially network‐based inference can drive the translation of systems biology to systems medicine by shifting the focus on the clinical impact.

3Moons is designed to deal with critical network aspects summarized as follows:

1. Integrative Networks: A probabilistic approach ‐ integrated with the next generation sequencing genome and transcriptome data ‐ to uncover cancer‐coordinated activity at modular level and determine change‐of‐state in proteins. Impacts: Preventive medicine and therapy through the analysis of protein pathways to identify candidate markers and possible targets for drug development.

2. Multiscale Networks: A spectral approach to overcome the limited resolution allowed by most algorithms designed to discover modularity according to functionally coordinated genes or proteins. Impacts: Elucidation of cancer signaling by examining temporal heterogeneity, i.e. permanent vs transient interactions due to slow (long lasting) vs fast (vanishing) dynamics, and propagation of fluctuations from small to increasingly larger scales.

3. Non‐extensive Networks: An entropy approach to deal with the complex multiomic dependencies and the related uncertainty arising from data non‐stationarities (pattern discontinuities or local clustering). Impacts: Identification of dysregulative and interactive cancer‐induced patterns through network‐adapted uncertainty and complexity measures such as non‐extensive ensemble entropies.

The budgeted cost for the project for the 2 years is a total of AED 398,000. There are many reasons for the success of the proposed project which include the collaboration of three very well established centers in the fields of bioinformatics and computational biology (CIT‐UAEU, CCS‐University of Miami and LISM‐CNR, Pisa Italy); The experience of the PI and the Co‐PIs in the field of cancer and bioinformatics; the availability of data, domain experts and excellent computational facilities at the College of Information Technology, UAEU.

The project has important significance to UAE. The visibility by the project will put UAE at the forefront of bioinformatics innovation, driving standardization of the technologies and methods. The project will create specific cancer markers database which UAEU will leverage in future projects with a vision to make significant contributions to medicine and healthcare.

While viruses cause significant proportion of cancers, their real impact in cancer etiology and complications is still unknown. Indeed, viruses may be implicated in cancer more frequently than previously considered, which may lead to rethinking cancer treatment and prevention strategies. Little effort so far has been exerted to screen for infectious agents associated with cancer. Our long-term goal is to determine how cancer treatment and prevention can be improved in virus-associated cancers. The objectives of the proposed research are to develop a computational pipeline to improve our technological ability to detect and identify new viruses associated with the disease; and to determine genetic variables associated with these viruses. Our working hypothesis is that there are previously unknown viruses associated with cancer that may induce an alternative etiology. We will test our working hypothesis on publicly available sequencing and clinical data. The rationale of this aim is that successful viral typing is critical for hypothesizing their causative role, and thus, the determination the future directions in cancer research, involving potential biomarkers, outcome prediction or tumor classification. Our central hypothesis will be tested by pursuing one specific aim: Identify new viruses potentially involved in human cancers. The proposed research focuses on an innovative approach, namely utilizing high-throughput sequencing data, to screen for all known viruses in various cancer types. This is the first study that will screen at least eighteen cancer types. Fifteen out of eighteen cancers were previously shown to have increased incidence in immunosuppressed patients or were linked to viral infection, hypothesizing virus causality. The data we want to use may provide valuable and timely return for the aim of the proposal. This contribution is significant because it is an essential step in the continuum of research that is expected to lead to new treatment strategies in patients with virus-positive tumors and better prevention through vaccine development and implementation procedures to prevent viral infection and the cancers they may promote. Once explicit information on cancer-associated viruses becomes available, regardless causal or not, virus-positivity may be used as biomarker for patients’ outcome prediction, early cancer detection, tumor localization and classification, as well as guide the type and intensity of therapy.

A new IBM-built Pegasus supercomputer came online this spring, offering UM researchers the ability to perform 160 trillion floating-point operations per second. Housed at the NAP of the Americas data center in Miami, it features a new generation of Intel central processing units, making it five times faster and more powerful than the previous Pegasus introduced by CCS about three years ago. Intel made the new state-of-the-art processing units, called Phi, available only to UM and a handful of other institutions.  Please visit eVeritas to read the entire article.

The system arrived earlier this year and promises to be more powerful than its predecessor.  It joins a family of systems making the environment one of the largest centralized academic cyber infrastructures in the country. The new Pegasus has been ranked at number 389 on the November 2012 Top 500 Supercomputer Sites list.  To learn more about Pegasus, visit the Pegasus Details Page page.

This is a beginner level HPC workshop where you will learn how to access and use your domain-specific applications on our CCS HPC resources that are available to you.

Please register on Event Brite for the sessions you wish to attend, and then complete the Survey Monkey questionnaire (See below) so that we can create a Pegasus account for you. Whether you have a Pegasus account or not, PLEASE NOTE that your registration will not be complete until you submit the Survey Monkey questionnaire (it’s only three questions!).

Please bring your own laptop.

Day 3 is Tuesday April 30th, five 2-hour breakout sessions between 9:30 – 5:30.
Max capacity 25 students per breakout session.
Pre-requisites are knowledge of nano, a Pegasus account, and a working knowledge of Unix as relevant to how CCS HPC resources work.
Breakout sessions will be held on CESM, WRF, HYCOM, and ROMS.

This is a beginner level HPC workshop where you will learn how to access and use your domain-specific applications on our CCS HPC resources that are available to you.

Please register on Event Brite for the sessions you wish to attend, and then complete the Survey Monkey questionnaire (See below) so that we can create a Pegasus account for you. Whether you have a Pegasus account or not, PLEASE NOTE that your registration will not be complete until you submit the Survey Monkey questionnaire (it’s only three questions!).

Please bring your own laptop.

Day 2 is Monday April 29th, 10:00 – 3:00 with a 1 hr lunch break in the middle.
Max capacity 25 students.
Pre-requisites are a working knowledge of nano and a Pegasus account. Day 2 will cover CCS resources and how they work (multiuser processes, queuing, batching, …), and further Unix as relevant to access of and ease of use of CCS HPC resources.

Breakout sessions will be held on CESM, WRF, HYCOM, and ROMS.

This is a beginner level HPC workshop where you will learn how to access and use your domain-specific applications on our CCS HPC resources that are available to you.

Please register on Event Brite for the sessions you wish to attend, and then complete the Survey Monkey questionnaire (See below) so that we can create a Pegasus account for you. Whether you have a Pegasus account or not, PLEASE NOTE that your registration will not be complete until you submit the Survey Monkey questionnaire (it’s only three questions!).

Please bring your own laptop. The workshop is over three days:

Day 1 is Thursday April 25th, 9:00 – 12:00.
Max capacity 25 students.
The only pre-requisite is for you to have a Pegasus account, which will be issued to you automatically at registration if you don’t already have one. Day 1 will cover absolute basics of Unix/Linux, VPN, Putty, and the GNU nano editor. You need to be familiar with nano by Monday.

Linux Basics
Linux Basics covers absolute basics of Unix/Linux, VPN, Putty, and the GNU nano editor. You need to be familiar with nano.

Linux Basics (Presentation slides)

Introduction to HPC
Pre-requisites is working knowledge of nano. Learn CCS resources and how they work (multiuser processes, queuing, batching, …), and further Unix as relevant to access of and ease of use of CCS HPC resources.

Introduction to HPC (Presentation slides)

Tools
Pre-requisites are knowledge of nano and working knowledge of Unix as relevant to how CCS HPC resources work.
CESM (Presentation slides)

WRF (Presentation slides)

HYCOM (Presentation slides)

ROMS (Presentation slides)

Click here to get a sneak peak, following this supercomputer from the warehouse where it was built to its arrival at the data center.