Tag Archives: cancer

Introducing Omics Logic

The Human Genome Project showed that while mankind’s genetic makeup is 99.1% identical, and only 0.9% of genetic variability creates vast variability that exists within the human species (Novelli 2010). Personalized medicine is the effort to prescribe the most appropriate drug for each individual patient based on their specific biology. Genetics explain some of the variations in responses seen during clinical trials. The variety of cancer mutations means that effective diagnosis and treatment of cancer must take into account a high degree of complexity. By sequencing individual cancer genomes, researchers and physicians may develop more targeted medical solutions. By using interpreted data in routine patient exams, clinicians can analyze how the symptoms of a disease in a patient correlate with their specific biology, resulting in a more effective treatment. Cancer is the second major cause of mortality in the United States, but targeted cancer therapies are bringing about an exponential increase in effectiveness over traditional cancer therapies.The potential exists to identify early indicators of disease, including cancer, in the form of biomarkers for early detection of a disease.


According to the Global Oncology Trend Report, global spending on cancer medications rose 10.3 percent in 2014, bringing the total to $100 billion, up from $75 billion in 2010. The rising cost of cancer treatment is linked to the emergence of precision therapeutics, which are costly to develop and often fail before they reach the market. While more effective, they target a smaller population which is hard to identify. The pharmaceutical industry recently turned to theoretical and computational modelling to improve the drug discovery process, lowering the cost of care in the process.


The cost of next-generation sequencing and other techniques that provide comprehensive whole-patient data is decreasing rapidly, making personalized multi-omics analysis increasingly cost effective and accessible. There is a long and costly effort to introduce precision treatment based on molecular data into both health delivery and pharma. A massive amount  of data is available, yet few know how to use the data insights effectively. To be truly effective, the data has to be analyzed effortlessly with an integrative approach. Using multi-omics data in developing an informed, personalized approach to treatment, access to effective clinical trials, and preventive strategies can provide major cost savings in terms of avoiding ineffective treatments, expensive diagnostic regimens, etc.


With affordable whole-patient scale data just around the corner, the challenge has now moved into the realm of extraction of meaningful insights from the data. To get the most value from multi-omics data analysis in clinical applications, Pine Biotech is developing an omics-first machine learning platform, OmicsLogic. The platform goes beyond analytics, integrating clinical knowledge with multi-omics raw data analysis for biomarker discovery and personalized molecular studies. As the field evolves and data continues to become available, algorithmic innovation is poised to be a driving force in solving healthcare ecosystem challenges. The wealth of data that is generated should be exploited – ultimately to improve care, and benefit consumers.




Novelli, G. Personalized genomic medicine. Int Emerg Med. 2010;5(Suppl 1):S81-90. doi:10.1007/s11739-010-0455-9

A Personalized, Precise Approach to Fighting Cancer

Cancer is the second major cause of mortality in the United States and targeted cancer therapies are bringing about an exponential increase in effectiveness over traditional cancer therapies. Traditional cancer treatment, such as chemotherapy, has come a long way in the past five decades, and care can be delivered comfortably, in an outpatient setting with manageable side-effects.

Screenshot 2017-06-13 16.29.07

Cancer is an important challenge for which personalized molecular medicine shows great promise. Recent advances in immunotherapy and genetic testing have been proposed to help transform care from one-size-fits-all to a highly specialized range of options that could be adapted to fit an individual’s molecular features. However, we are still far from understanding and navigating cancer.


The variety of cancer mutations means that effective diagnosis and treatment of cancer must take an individual approach. By sequencing individual cancer genomes, researchers and physicians may develop more effective treatment solutions. A multi-omics approach based on big data analyses could lead to substantial advances in cancer treatment, ushering in an exciting new paradigm in cancer treatment.


Accessibility of new technologies such as next-generation sequencing remains a barrier for many patients. The skills needed, and associated costs of  technology have prevented physicians and patients from using precision oncology to its full potential so far. The skills and equipment required to collect and analyze genomic data can be expensive. And generally, Bioinformaticians have to go through rigorous training in linux, python, R, databases, visualization and an endless list of new algorithms to even understand the complex datasets.


To encourage more researchers to make use of the potential and power afforded by big data, we designed a free series of lessons which simplify the computational aspects of analysis to allow scientists from all backgrounds to move forward in the world of big data. The cancer series has practical, hands-on projects that allow students to practice analyses with data adapted from real datasets.



The course will cover some of the important aspects of breast cancer:

  1. molecular indicators of cancer: deregulation of cellular checks and balances, uncontrollable growth, alternate signaling, altered immune responses, etc.
  2. Factors that contribute to cancer heterogeneity – “levels of biological regulation”
  3. Response to treatment studied with cancer cell lines.
  4. Beyond cell lines – PDX models: the role of microenvironment and the use of animals in research.
  5. TCGA data – real patients: miRNA-seq, RNA-seq, Exome-seq and clinical data.
  6. Deeper look into clinical data: combinations of treatments, many ways to diagnose cancer, why molecular data is critical.
  7. Future of Cancer: New treatments and findings that could change current cancer treatments.

Exponential growth of data in biomedicine needs a skilled and experienced workforce

A wealth of public domain biomedical data is available to anyone online, however, data analysis technology can only be utilized by those trained to work with it, leaving researchers without the proper skills and tools drowning in the overwhelming amount of information. It might sound intimidating, but in reality, this is GOOD NEWS!

Data science is a fast-growing branch of the technology landscape, and with IT and analytic skills highly sought after by employers, data scientists pull in hefty salaries for their expertise. There is an urgent need for quick, cost-effective, and accurate data analysis in the field of biomedical research as well. This need is driven by the continually expanding wealth of patient data, coupled with the need to synthesize and understand the data’s importance in medicine.  One of the most exciting areas where new discoveries are being made all the time is in the molecular data field. The datasets there are huge, but every new project brings us closer to understanding complex diseases, evolution and human longevity.

Today, biomedical data is typically analyzed by a trained bioinformatician – a tech-era mash-up of “biologist” and “statistician” who breaks down and interprets large sets of medical data in a clinical or research setting. But for the most part, bioinformaticians have to go through rigorous training in linux, python, R, databases, visualization and an endless list of new algorithms that are complex and require deep understanding to use. Just to get to a meaningful project might take years depending on how well you are versed in all of these technologies. All of these skills essentially turn you into a computer scientist first and then place you under a biologist’s or clinicians’ authority to guide “real research”.

But more and more biologists, clinicians and even patients want to play a role in these discoveries. At some point, you had to know Fortran to use a computer – that is until a mac came out or Windows with their visual interfaces making it a household item. And then kids could play with the infinite scripts (games) that leveraged this complex technology. Well, that’s exactly what we would like to do – let’s make bioinformatics easy, visual and intuitive. We think we can do it together.

To start, we’ve put together a number of projects that one can analyze using our visual bioinformatics platform. Here are a few examples:

“PDX Models: Tumor-stroma interaction” inspired by the publication, Whole transcriptome profiling of patient-derived xenograft models as a tool to identify both tumors and stromal specific biomarkers by Dr. James Bradford. The project’s approach focused on comparing several different breast cancer types using RNA-seq and machine learning methods including 79 PDX mouse models with human primary tumors. PDX models maintain more similarities to the parental tumors than a traditional cell line does. Subsequently, alternative analysis of experimental data provided deeper insight into the problem and identified new biologically meaningful group-wise associations between tumor and stroma genes.

“Cell Lines: multi-omics network of associations” rests on the publication of Expression Profiling of Macrophages Reveals Multiple Populations with Distinct Biological Roles in an Immunocompetent Orthotopic Model of Lung Cancer. The approach focused on the molecular features associated with responses of a collection of 70 breast cancer cell lines to 90 experimental or approved therapeutic agents. This project included both RNA-seq, mutations, and IC50 drug values. During analysis, biassociation was utilized with the expression and mutation variant results as well as drugs (GI50) to find relationships between the datasets.

Explore these datasets online, and learn more about our bioinformatics platform T-BioInfo, at http://edu.t-bio.info/projects/