SEATTLE: A biomedical facility is hoping that data mining can do for breast cancer and heart disease research what it did for retail giants such as Wal-Mart.
The Windber Research Institute said it has chosen Teradata to help the facility create the first central data warehouse for molecular and clinical information by combining the information of five organizations that generate data ranging from from protein interactions to metabolic pathways. The partners announced their plans at Teradata’s Partners conference taking place here this week.
“”We felt the people who could do what they did for the Wal-Mart computer systems (uniting data from multiple sources) would certainly be able to do what we want to be able to do at the Windber centre,”” said Nicholas Jacobs, president and CEO of the institute, based in Windber, Pa. “”We will be creating what has been dreamed about for many years – an integrated biomedical warehouse to help find the cause of cancer.””
The institute uses data from several sources and studies the relationship between proteins, genes and disease. If it is determined that a woman who goes to a clinic for a mammogram has a suspicious lesion, a biopsy will be done and the lesion is removed. It is then sent to Windber to determine what parts will be studied and analysed and then entered into the data repository.
“”Each sample provides 450 fields of information about the person’s health or lifestyle such as whether they smoke, whether they grew up in an urban or rural area or whether they drank three glasses of wine a day. We will be able to then put the demographic and molecular data together with clinical data,”” Jacobs explained.
For each tissue sample, 166MB of information will be created. Physicians will hopefully be able to query the database to determine the on/off switch for the cancer.
Research at Windber will be used as the basis for vaccine being used in a trial to stop the recurrence and spread of breast and other forms of cancer. The goal is to develop individualized medicine to patients by identifying patient-specific causes at the molecular level. The hope is the vaccine will eliminate the need for surgery and treatments in the future.
“”We believe we have created, under one roof a completely integrated system where molecular and demographic data can be collected and tracked from six months to 16 years ago and look at where the markers of malignancy begin to appear,”” said Jacobs.
To date, Teradata is not working with any Canadian cancer researchers on similar projects, but Martin Hill, vice-president of business development at NCR’s government systems group, said there is interest in this kind of work in Canada.
It is not the first time Teradata has worked in the biomedical arena. Scientists at the Salk Institute in San Diego, Calif. have been using data mining to understand the growth of brain cancers in mice since 2002. Instead of taking a year to study one gene, researchers say data mining allows them to study the potential interactions of 13,000 mice genes in one week.
Comment: [email protected]
