Featured Article
From the start, a potentially new medicine is unlikely to be a hit. Biotechnology and pharmaceutical companies try to improve that ‘batting average’ with any ’omics that can be thrown at the problem. This includes genomics, proteomics, lipidomics, and metabolomics. Moreover, the various kinds of ’omics can be applied to drug discovery and development in many ways.
Before exploring ’omics options for improving the batting average of making new medicines, let’s figure out what number medicine makers face. The answer is: It depends. It depends on who is asked or the study examined. It depends on the class of medicine, like cardiac or cancer drugs. Over the years, a number of around 10% seemed to be a good guess. Recently, MIT economist Andrew Lo and his colleagues compiled a large dataset and concluded: “13.8% of all drug development programs eventually lead to approval ….”1 Keep in mind, that’s the average across diseases, and the odds get worse for some diseases, and better for others. For instance, cancer creates the worst batting average—3.4%; vaccines against infectious diseases turn out to be the most likely to succeed—33.4%. At any batting average, though, clinical scientists search for ways to do better with ’omics.
“’Omics can be used in many stages of drug discovery and development,” says Steve Fischer—market director, academia & government/life science research segment, Agilent Technologies (Santa Clara, CA). “The power of ’omics comes from the study of many genes, proteins, lipids, or metabolites at the same time, with as little bias as possible.”
In fact, Jim Russell at the Centre for Heart Lung Innovation at St. Paul’s Hospital (Vancouver, Canada) and his colleagues believe that combining various forms of ’omics could be used to develop new treatments for sepsis. As Russell and his coauthors noted: “Treatment is complicated because sepsis is heterogeneous, explaining the lack of effective drugs.”2 They added: “Multi-’omics confirmation refines mechanistic understanding so that high probability drug targets can be identified.”
Some of the most common uses of genomics in clinical research involve cancer. For example, Gokham Yildiz of Karadeniz Technical University (Trabzon, Turkey) studied treatments of hepatocellular carcinoma (HCC)—the most common kind of liver cancer in the world—with genomic and transcriptomic tools.3 Yildiz reported that “the results of high-throughput drug treatment experiments on HCC cells analyzed in the present study indicate that molecular-targeted, personalized chemotherapeutic approaches should be developed for the treatment of HCC, since distinct HCC cell types respond differently to the same drug treatments.” Similar conclusions can be drawn for many kinds of cancer.
These examples show some of the range of diseases that could be targeted with ’omics-related data. Now, let’s see how scientists can do even more.
DNA and the resulting proteins can help scientists learn more about diseases, including where and how to hit them. (Image courtesy of Agilent Technologies.)Ups and downs
The best kind of ’omics to use depends on the stage of making a new medicine. “In discovery, proteomics or genomics can be used to determine drug targets,” says Fischer. Usually, the target is a protein, occasionally a gene.
“In the development phase, metabolomics can be used to study the metabolic pathways of a drug, to optimize the molecule for efficacy and minimize off-target effects, including toxic or reactive metabolic products,” Fischer explains. “In clinical trials, genomics can be used to classify patient phenotypes that will respond to the therapy, have favorable metabolism or have toxicity issues.” To track a medicine’s efficacy or look for toxicity markers, clinical scientists can use metabolomics.
Despite all of the advances in ’omics-related tools, that’s not enough to make it instantly easier to create new medicines. “Each ’omic has its own challenges, but as a group ’omics have the same primary issue, that they are dependent on well-designed experiments using large numbers of samples to correlate disease to biochemistry,” Fischer explains. “These experiments generate large amounts of data that then require analysis and statistics utilizing sophisticated software to determine relationships.”
The question is: What do those relationships mean? “While ’omics are tools to develop an understanding of a disease, correlation is not causation,” Fischer cautions. “Therefore, any results of ’omics experiments require substantial additional work to translate the correlation to biological understanding.”
To build that understanding, scientists need tools that dig deeper into the questions at hand. From Agilent, Fischer says that the 1290 Infinity II LC system combined with the 6546 LC/Q-TOF system with Lipid Annotator software is its latest improvement in applying ’omics technology to drug discovery and development. He says that this system is “designed to assist researchers with analysis of complex lipid samples, deliver outstanding separation and mass measurements, as well as provide a greater level of confidence through the simultaneous delivery of wide dynamic range, high-resolution, and mass accuracy.” He adds that the “associated Lipid Annotator software provides fast analysis of MS/MS spectra produced by the 1290/6546 LC/MS systems to annotate the observed spectra with a low false positive rate, which provides high confidence in the analytical results, allowing researchers to more quickly understand the biology being studied.”
The batting average for making new medicines is challenging in most cases, but various forms of ’omics could improve this statistic. (Image courtesy of the author.)Home-run hopes
Part of the strategy for getting more hits from drug candidates depends on getting to the plate more. In part, that can be accomplished by getting more of the world involved in the game. To do that, as many people as possible need access to the best tools.
Expanding the use of ’omics can arise from affordable platforms. So, Illumina (San Diego, CA) developed its iSeq 100 Sequencing System, which is described as a next-generation sequencing (NGS) system that “delivers exceptional data accuracy, at a low capital cost, making Illumina technology available to virtually any lab.” The price is under $20,000.
All ’omics platforms need powerful software to get the most out of the mountains of data generated in searching for new medicines and trying to develop them. In Houston, Genialis provides a software platform powered by artificial intelligence and is designed for drug discovery. Genialis points out that its software “supports the analysis of high-throughput genomic, transcriptomic, and epigenomic sequencing data to accelerate your drug discovery efforts.”
Regardless of what technology is used, Genialis software can work with the results. As the company notes, its software ensures “apples-to-apples interpretation of your NGS data, no matter if it’s public, proprietary, historic, or newly acquired. With validated informatics pipelines, and built-in normalization and QC/QA workflows, you can confidently, consistently interpret your data.” When needed, Genialis even provides data scientists to help customers mine and analyze information.
As clinical scientists collect more data and find new ways to explore it, the batting average in getting new drugs to market could improve. In any case, the baseball great Babe Ruth once said, “Every strike brings me closer to the next home run.” Maybe at the very least, exploring all of the ’omics options will get scientists to the new-medicine plate more often, and the home runs will start increasing.
References
- Wong, C.H.; Siah, K.W. et al. Estimation of clinical trial success rates and related parameters. Biostatistics 2019, 20(2), 273–86; doi: 10.1093/biostatistics/kxy072.
- Russell, J.A.; Spronk, P. et al. Using multiple ’omics strategies for novel therapies in sepsis. Intensive Care Med. 2018, 44, 509–11; doi: doi.org/10.1007/s00134-018-5122-z.
- Yildiz, G. Integrated multi‑omics data analysis identifying novel drug sensitivity‑associated molecular targets of hepatocellular carcinoma cells. Onocology Letts. 2018, 16, 113–22; doi: 10.3892/ol.2018.8634.
Mike May is a freelance writer and editor living in Texas. He can be reached at [email protected]