Expression Analysis is partnering with biotechnology firm HemoShear to generate and analyze RNA information from several thousand human samples that will populate a drug database that HemoShear is developing.
HemoShear will use the information generated by EA to profile how human vascular cells respond at the genomic level to about 75 drug compounds that span a wide range of drug classes and have been accepted, black-boxed, or withdrawn from the market, the partners said.
HemoShear intends to market the vascular pharmacology database to pharmaceutical companies, who will be able to use it "to establish a true risk profile of their compounds and investigate potential positive or negative effects,” Nicole Hastings, HemoShear's vice president of laboratory operations, said in a statement.
For example, “a number of diabetes drugs … have been shown to have adverse cardiovascular effects in humans. We can provide insights about the risks associated with continuing development of new compound candidates by comparing to other drugs in our database that are related by class, genomic signature or mechanism of action," Hastings said.
This project is funded by a $4.3 million Phase II Small Business Innovation Research grant to HemoShear from the National Heart, Lung and Blood Institute.
Wendell Jones, EA’s vice president of statistics and bioinformatics, told BioInform that his firm will have a two-pronged role in the partnership.
He said that EA has begun using next-generation sequencing technologies to generate transcriptome data from more than 2,000 human RNA samples that HemoShear will use to evaluate the efficacy and safety of drug targets.
Transcriptome sequencing will enable HemoShear to develop the “most comprehensive and sustainable database for evaluating drug vascular safety and efficacy” because it can "reveal the expressed quantities of protein-coding messages and isoforms of all active genes as well as detect novel post-transcriptional modifications,” resulting in a comprehensive database, Jones said in a statement.
In contrast, traditional techniques such as microarrays, which EA also offers in its services portfolio “can only detect changes in expression of predetermined genetic content within a more limited dynamic range,” he added.
The next step in the partnership will be to run the sequence data through EA’s RNA-seq analysis pipelines, Jones told BioInform.
The company intends to provide the bioinformatics tools and computational infrastructure needed to process HemoShear’s genomic data in a defined format with consistency and speed.
As part of the analysis, EA will work with HemoShear over the next few months to tailor its RNA analysis pipeline to meet its partner's needs, Jones said.
EA's current pipeline includes tools for quality control and to filter sequences so that extraneous reads — for example those that include artifacts from the amplification process — can be removed prior to analysis, he said.
It also includes tools for quantifying transcription at the gene, transcript, and isoform level as well as methods of converting sequence data into the cel file format used by Affymetrix microarrays, he said.
This way, customers who have internal informatics pipelines that accept the Affymetrix file format but who would like to use RNA sequencing to generate data in their projects can still use their existing infrastructure for the analysis, he said.
In addition, the company has made an effort to create a processing pipeline that allows clients like HemoShear to “process several hundred or even thousands of samples and get very consistent output,” Jones said.
EA can also perform more in-depth analysis of customer data. For instance, it has tools to detect variants and fusion events, though these more specialized analyses are offered on a case-by-base basis, Jones said.
Meanwhile, HemoShear’s internal bioinformatics team is working on a pipeline that the company will use to “streamline the identification of regulated genes” and perform pathway analysis once it gets data back from EA as well as to find correlations between the transcriptome of each sample and the specific drug targets in the study, Rob Figler, HemoShear’s scientific director, told BioInform.
“The idea is to essentially provide a predictive set of transciptome fingerprints for the pharmaceutical industry to use to identify and mitigate risk in the drug development process,” he said.
Figler declined to provide specific details about HemoShear’s internal bioinformatics pipeline because parts of it are proprietary, he said.