Reducing the ‘what ifs’ in drug discovery
With emergence of specialised software applications, drug discovery has become a highly cost-competitive area for Indian pharma companies. Nagesh Joshi examines the use of specialised software applications in drug discovery
Drug discovery was the main aim of any pharma company, prior to the advent of the doctrine that companies could have a profitable business model without selling a drug they actually ‘invented’. A pharma company could just make changes in the ‘process’ and have a ‘generic’ version of a drug. This doctrine was supported by most of the developing economies in order to protect their populations from the over-pricing of the patented ‘original’ versions of various life saving drugs.
After the advent of WTO norms, which have been accepted by almost all nations now, product patents on original drugs have become recognisable even in developing nations. Companies have been forced to wait until the patents lapse to market generic versions.
The drug discovery process has become more and more complex, time consuming and very expensive, causing a many-fold increase in the R&D budgets of pharma companies. It still remains the best chance to make money for a pharma company, but has become unaffordable for all except the so-called ‘big pharma’. However, the other leaner business models are emerging. One case-in-point being the recent new drug development agreement between Nicholas Piramal India Limited (NPIL) and Eli Lilly, wherein NPIL will develop, and in certain regions, commercialise a select group of Lilly’s pre-clinical drug candidates.
Biopharmaceutical companies are also coming up with cheaper, faster and more efficient ways of getting to new chemical entities. The advent of in-silico technologies for optimising the R&D pipeline from basic biology phase to chemistry phase, to lead optimisation and so on up to clinical trials has also led to considerable improvements in efficiency.
“Another business model, especially important to India, is to spin off R&D as a separate business entity to raise resources, as well as reduce risk. Variants of this strategy have been followed by Ranbaxy, Dr Reddy’s Laboratories, NPIL—the three largest pharma companies of India—for high-rewards in the area of research and development,” says Dr Vijay Chandru, Chairman, Co-Founder & Chief Executive Officer, Strand Life Sciences.
An important trend here is the rising investment in health care related expenses on IT by many developing nations, including India, that are opening new doors.
Need for computing software
The traditional method of drug discovery, as known to all pharma companies and research scientists, is a highly serendipitous process. Therefore, the cost of developing a successful new molecule also reflects the expense of failed molecules. Thus, the scientists/researchers are always looking for ways to avoid failures and to improve their chances of success.
Therefore, certain technologies, which facilitate the enhancement of predictability, for example, computer aided drug design (CADD) or molecular modelling, are finding increased acceptance in the process of drug discovery. Most innovation driven research companies are utilising CADD as a fundamental step in optimising their research activities and finding ways to arrest the failures earlier. There are computing softwares which help knowledgeable scientists in the ‘what-if analysis’ by studying various molecule-protein interaction scenarios, comprehensive exploration of the chemical and biological space without actually making them, design better leads, detect problems at molecular level at an early stage so that time and effort in the essential experimental work in the laboratory is optimised, thus improving overall research productivity.
There are two main challenges that the drug discovery domain is facing presently:
1. The rising cost of the process of drug discovery itself, with scarce talented resources and rising input costs affects the efficiency of the process
2. The intellectual property rights (IPR) protection issues arising because different countries follow different norms affects the effectiveness of the process
Apart from these two, there are other nagging issues such as, the limited success pharma and biotech companies have achieved in terms of reducing the development time period, in spite of the availability of several reliable in-silico methods and technologies.
The rising number of generic companies as well as ‘one product’ or ‘one technology’ companies are reducing the market share enjoyed earlier by the major pharma companies, putting pressure on their bottom lines ,as well as top lines.
The present scenario
While the technologies have not matured to the extent that their output is always right, technology products, as a tool in the hands of a knowledgeable scientist, is a significant contributor towards improving research productivity. Therefore, the expectations from technology are increasing day by day.
“Amongst the few technology providers in CADD and molecular modelling domain, companies which are innovative and are keeping pace with the evolving science are likely to survive and grow rapidly. On the other hand, significant opportunities for students are emerging in the CADD area, as it is increasingly adopted as a fundamental activity in most drug discovery programs globally,” says Atul Aslekar, Chief Executive Officer, VLife Sciences.
“A typical research program consists of two distinct phases—discovery and development. In the first phase, CADD is increasingly used as a starting point”, says Dr Sudhir Kulkarni, Principle Scientist at VLife Sciences.
CADD provides a strong tool to scientists, which enables them to custom design a new molecule, keeping in mind the specific requirements of protein causing disease condition. It also helps scientists to try out various ideas in a short time, as compared to conventional methods. In-silico technologies like CADD enhance the exploration space for a new molecule. Novel virtual screening technologies are enabling scanning of the chemical possibilities on variety of criteria such as ligand binding, absorption, distribution, metabolism, and excretion (ADME) properties, etc. CADD technologies are helping in understanding drug-target interactions at a molecular level, which helps in designing better drug candidates. In the hands of an able scientist, CADD can not only significantly save the invested time, but can also lead to higher quality of pre-clinical candidates with higher probability of success, in later investigations.
Different research organisations, trade magazines and industrial bodies have put the research expenses going into drug discovery anywhere between $500 million-1.2 billion. However, an expenditure of about $900 million-1 billion may be considered as a reliable estimate from the amount of R&D expenses disclosed by all the big pharma companies, and the number of new drugs they have been able to discover over the last decade.
Anu Acharya, Chief Executive Officer, Ocimum BioSolutions, places the potential size of the drug discovery software market as $2 billion. According to her, “The drug discovery software market in India is at a nascent to mid-maturity stage.”
An estimate of the failure rate could be had from the reality, that of the approximately 5,000 compounds that enter the medicinal chemistry and drug metabolism and pharmaco-kinetics (DMPK) evaluation phases of drug discovery, only one succeeds and becomes a drug.
“There are several pain points that specialised software tools can help relieve for scientists working on drug discovery. Specialised software can either be used to manage data and analyse it or to generate very large amounts of data by carrying out experiments on a scale hitherto impossible,” informs Dr Chandru of Strand Life Sciences.
The software applications used for generation of data are usually in the preliminary stages of the drug discovery process. These stages involve basic biological and chemistry research for identifying targets, biomarkers, genes responsible for the disease etc. on the biology side. On the chemistry side, it involves a lot of high throughput screening processes to quickly and cheaply eliminate potentially less useful hits. Software tools used during this stage run specialised algorithms and applications for identifying patterns, outliers and specific features in data points generated through experiments. Some applications, such as the embedded software in various gene expression analysis equipment, help in generation of such data points.
In the later stages of the process, data management and analysis for better and more efficient decision support become more important. The software applications used here are focused more on statistical data analysis and modelling ,using various machine learning-based techniques.
The main steps in which software applications prove helpful are QSAR modeling, computational chemistry modelling for early ADME-Tox and DMPK predictions. Recently, data at the stage of clinical trials has also been put to statistical tests using high-end statistical analysis software tools.
|Virtual screening: Pharmaceutical companies are always searching for new leads to develop into drug compounds. One search method is virtual screening. Here, a large chemical space is screened against a protein to shortlist those molecule, which may have better binding affinity for the protein. If there is a “hit” with a particular compound, it can be extracted from the database for further in-silico testing and then taken into the laboratory for physical validation of the in-silico hypothesis. With today’s computational resources, several million compounds can be screened in a few days on sufficiently large clustered computers. Pursuing a handful of promising leads for further development can save researchers considerable time and expense. ZINC database is a good example of a virtual compound library.
Sequence analysis: In CADD research, one can study the genetic sequences or the amino acid sequences of proteins from several species. It is very useful to determine the similarities or dissimilarities based on gene or protein sequences. With this information one can infer the relationships, search for similar sequences in bioinformatic databases. There are many sequence analysis tools that can be used to determine the level of sequence similarity.
Homology modeling: Another common challenge in CADD research is determining the 3-D structure of proteins. Most drug targets are proteins, so it’s important to know their 3-D structure in detail. Human body has several hundred thousand proteins. However, the 3-D structure is known for only a small fraction of these. Homology modeling is one method used to predict the protein 3-D structure. If the structure of a specific protein (target) is not known, then it is modeled, based on the known 3-D structures of proteins (templates), sequentially similar to the target, using the homology modeling technique.
Quantitative structure activity relationship (QSAR): QSAR is the process by which chemical structures are quantitatively correlated for their biological activity or chemical reactivity, based on well-defined statistical modeling process. The correlations and the statistical models are then used to predict the biological response of the other chemically similar structures.
Drug lead optimisation: When a promising lead candidate has been found in a drug discovery program, the next step is to optimise the structure and properties of the potential drug. This usually involves a series of modifications to the primary structure (scaffold) of the compound. This process can be enhanced using software tools that explore related compounds with respect to the lead candidate.
Similarity searches: A common activity in drug discovery is the search for similar chemical compounds. There are variety of methods used in these searches, including sequence similarity, 2D and 3D shape similarity, substructure similarity, electrostatic similarity and others. Several chemoinformatics tools and search engines are available for this work.
Pharmacophore modelling: Pharmacophore is defined as the three-dimensional arrangement of atoms, or groups of atoms, responsible for the biological activity of a drug molecule. Pharmacophore models are constructed, based on compounds of known biological activity and are refined as more data are acquired in an iterative process. The models can be used for optimising a series of known ligands or, alternatively, they can be used to search molecular databases in order to find new structural classes.
Drug bioavailability and bioactivity: Many drug candidates fail in Phase III clinical trials after many years of research and millions of dollars have been spent on them. And most fail because of toxicity or problems with metabolism. The key characteristics for drugs are absorption, distribution, metabolism, excretion, toxicity (ADMET) and efficacy—in other ords—bioavailability and bioactivity. Although, these properties are usually measured in the lab, they can also be predicted in advance with bioinformatics software.
Courtesy: VLife Sciences
Quality of the software suite
Reliability and predictability of performance, consistent delivery and accuracy of output, equal ease-of-use for beginner, moderate and advance skilled users, and flexibility of analysis/performance options for users are few important qualities of good software. A vendor should ideally, have high quality resources for developing the software with rich experience in having actually done the laboratory experimentation that the software is going to aid in, have quality development, data security and testing processes in place, rapid and end-to-end customer support capabilities in case of queries and/or failures of any scale and type.
Phases in drug discovery that can use software:
The following stages require software applications to support efficient decision making at each of these stages. They are arranged in the order of appearance in the drug discovery pipeline:
1. Systems biology modelling 2. SNP & gene expression analysis 3. Biomarkers 4. Pathway analysis 5. Molecular profiling 6. Computational chemistry 7. Focused libraries 8. QSAR modeling 9. Lead optimization
Courtesy: Strand Life Sciences
The software products used in drug discovery domain are priced differentially. Pricing is highly flexible as the deliverables are quite readily customisable. Most vendors prefer enterprise-wide licensing deals with annual maintenance contracts, since they usually have lock in periods (commonly three years).
The more advanced or specialised products are still sold on outright purchase basis. These are typically for very specialised and/or limited access use. Drug development agreements are on the rise and industry analysts predict many more pharma companies will follow the model set by the NPIL-Lilly deal. The GVK BIO Wyeth Hyderabad Chemistry Center, a built-to-suit research centre for Wyeth Pharmaceuticals located in Hyderabad, is another example.
In conclusion, though the market for drug discovery/development software products is still at a fairly nascent phase in India, it seems set to grow as Indian pharma companies position themselves as partners in drug discovery and developers. Companies like Strand Lifesciences, Ocimum BioSolutions, VLife Sciences and the likes will reap the benefits of being the early birds in a sunrise industry.