Chem4IR

One of the most prominent resources of the fourth industrial revolution will be Data Science. There is already a computational drug discovery industry worth billions of dollars. Currently Bangladesh has no established specialized chemical research and/or academic institute that can perform research on chemical data using Artificial Intelligence. Because of the unavailability of such resources it becomes uncertain to have skilled generation to lead the scientific fields during the fourth industrial revolution. Therefore, we plan to establish a specialized research laboratory which will perform computational research and implement them using laboratory-based experiments. This platform will be used to train up younger professionals to learn data science and utilize it to govern 4th industrial revolution. These trained personnel will be able to compete with the international drug discovery market and provide a lifeline for the future of the pharmaceutical industry of the country.

Drug discovery, a multibillion-dollar industry, is contributing continuously to save human lives. Industries around the world are contributing to this sector while Bangladesh could not contribute significantly though we have a definitive prospect to contribute soon. On the other hand imported API based pharmaceutical industries had been smartly contributing to the global  sectors. Output has grown a thousand times since 1982, to US$2 billion, around 1% of gross domestic product, making it the biggest white-collar employer in Bangladesh. The industry supplies almost the entire domestic market and more than 100 other countries.
Bangladesh has been scheduled to leave the LDC category in 2024. Hence, the industries of our country will no longer have the access to a special World Trade Organization (WTO) waiver which exempts the industries from the Agreement on Trade-Related Aspects of International Property Rights (TRIPS). These markets can make whatever drug they want, drastically cutting costs and improving availability, the most obvious benefit of the waiver. Without this exemption the cost of production will dramatically increase due to patent-related expenditure. Simultaneously, the cost of acquisition of drugs will increase exponentially, which will put the lifesaving medicines out of the reach of mid to low income consumers in Bangladesh.

In a broader sense, this project would serve the entire pharmaceutical industry, starting from a consumer level at the bottom end of the logistical chain, to the professionals who would be working in the industry at the top end of the pyramid. The necessity of this project cannot be stated enough, as without the patent exemption and a subsequent work force with an enriched knowledge base to fill this void, the entire industry risks failing. At the very least, the country would experience a spike in lawsuits from foreign organizations if we do not follow the proper patenting methodologies needed to survive in a globally competitive market, where we no longer have the element of sympathy. The direct impact of this would be our impoverished classes, who without affordable medicate, could produce a public health crisis.
For our young and even experienced professionals, who are entering a global market without an adequate knowledge base, they risk becoming either uncompetitive or irrelevant. Hence, the beneficiaries of this project would be our entire healthcare industry, the populace that feeds off it to survive and professionals who are direly needed to be internationally competitive in the market of the future.

There is an absence of visible resources or expertise in the cutting-edge machine learning driven field of drug discovery in Bangladesh. We do not have any dedicated schools, neither educational nor production based facilities, thus leaving a void in the knowledge base in a critical field. We want to create a platform that will provide the skills to develop advanced methods in drug discovery which in collaboration with national and international chemical and pharmaceutical companies will lead to patents and discoveries of drugs and industrially important organic molecules. A platform where we, Bangladeshis, will be able learn and implement their learning at the same time. 4IR technologies are not spatial in nature, so anyone with the skills can work as a freelancer for any company around the world. Chem4IR will partner with likeminded companies around the world so that they can hire people from our institution and collaborate with us so that our graduates can help them in their projects and get paid. Through building up the talent pool and collaborating with labs and companies around the world, Chem4IR will produce the patents, software tools that will save the country a fortune in the longer run, as we will not be depended on buying patents from other companies or countries anymore. With the imminent graduation from LDC, we must mitigate the effects of the sudden loss of access to a special waiver from WTO exempts our country from TRIPS. (Agreement on Trade-Related Aspects of International Property Rights). We will be using open source dynamic molecular and protein engineering simulation tools such as GROMACS, PyMOL, Coot, MODELLER, SWISS PDB, Discovery Studio etc. We also have used tensorflow, keras, scikit learn and other Machine learning packages in our previous projects where we have used public repository data to analyze breast and brain cancer image models. Such tools shall be used to intricately analyze the enormous chemical space which is still largely unexplored. In addition, we will try to establish new tools to make these methodologies more efficient and cost effective. We want to create the stage to motivate the youth in our country that the new world of technology always waiting for them. We want to ignite the fire within them and gradually set the stage when we will be able to spread the product in the market, out of our research lab.

Bangladesh potentially leaves the LDC category in 2024. It’ll no longer have access to a special World Trade Organisation (WTO) waiver which exempts the industry from the TRIPS agreement. 

Most of the hundred pharma companies operating in Bangladesh make so-called generics, or non-branded medicines, the patents of which have often expired. Around a fifth of drugs produced in the country are patented in other countries, something which is made possible by the waiver which until 2033 allows LDCs to produce patented drugs without first asking patent holders.

The most obvious benefit of the waiver is that companies can make whatever drug they want, drastically cutting costs and improving availability.  

And now that the waiver is soon going to expire, no company in the country has yet to propose a solution to mitigate the effects of losing the waiver. The solution is to adopt a policy for a machine learning driven computer aided drug discovery lab. Without such a lab, the companies will have to buy patents for the medicines they want to produce. If they had their own labs by now, they could have produced patents for a lot of drugs and wouldn't have to rely on the special waiver, in the first place. 

As no other company in the country is thinking about dealing with this implosion with the help of the tools needed in the upcoming age of fourth industrial revolution, we hope to provide a lifeline to the pharma industry from imminent catastrophe. 

With our current undergraduate curriculum, we can't make a giant leap. In the very beginning our goal would be to develop skilled manpower in the field. We plan to mentor senior and final year undergrad students studying chemical, pharmaceutical and life sciences so that they can pursue higher study on applying machine learning/artificial intelligence in these branches of science at reputed universities around the world. This will help us in creating the talent pool and increasing our network and collaboration.  In Chem4IR we have already designed our curriculum in such a way that a freshman in chemistry can take our series of courses and can eventually develop his skills interlinking chemistry and data science. But now comes another challenge, that is the lack of employment or working opportunity. That's why we will collaborate with nationwide research institutes and present them our solutions and make collaborations with them so that people who will graduate from our courses that get involved in collaborative research projects. We will also motivate the leaders in pharmaceutical industry to undertake computational drug discovery research projects and make employment opportunity for people who are gaining skills in this field. Chem4IR will partner with likeminded companies around the world so that they can hire people from our institution and collaborate with us. Simultaneously, we will also have our own research projects to develop local healthcare solutions e.g. developing low cost diagnostic solutions for different types of cancer using machine learning algorithms. 

We haven't yet started serving people as we are in the concept stage right now.

In the first year we plan to serve 200 students and young professionals and make them interested to develop their skills in this emerging field and show the right path to evolve.

In 5 years: 1500 (estimated)

1. Lack of consciousness within the industry regarding the imminent situation.
2. Inadequate understanding about data sciences and analytics methods and tools.
3. Inadequacy of workspaces and tools.
4. Insufficient mentoring due to the field’s infancy in nature. 

With consistent and sustained funding from donations, awards, grants and collaboration with national and international chemical and pharmaceutical companies. Additionally, an interest in computer processed analytics and data sciences can be generated through promotion of these technologies, which is adequately taking place already. In time, a dedicated and knowledgeable workforce in these fields will be built and that workforce can be our allies in achieving our task to the fullest sense. Campus based seminars and advertisement of these emerging fields can draw non-disciplinary pupils into our cause, which can prove to be a prudent force when the need arises. Furthermore, we would ideally seek help from our university tutors and teaching staff, who have more insight into where the industry, and can provide us with guidance on how we should proceed with a project like this. Finally, we must be cautious of the fact that Pharmaceutical industries in Bangladesh are one of the most regulated industries in the country, hence a collaboration with government officials and state-interest groups would be highly wise and beneficial. 

As the solution is in conceptual phase, only THREE of us are currently working on this project.

We already have worked on a mathematical model of cancer incidence, where we have proposed our own theory regarding the matter. 

We also have worked on protein binding and solvation simulations, drug discovery, protein classification using artificial intelligence and other web-based tools.  

We already have published five research articles in reputed research journals and two more will be submitted for review, shortly. 

We are also proficient in using open source dynamic molecular and protein engineering simulation tools such as GROMACS, PyMOL, SWISS PDB, Discovery Studio etc. We also have used tensorflow, keras, scikit learn in our previous projects where we have used public repository data to analyze breast and brain cancer image models.  We also have an understanding of API based web solutions, in which we can use the Application Processing Interface of a third-party system to generate data on our platform. In the technology world, APIs are being regarded as the most important web-solution of the coming world. 

We are currently partnering with Red-Green research center Bangladesh to analyse protein structure and run dynamical protein and drug binding and solvation simulations to understand and contribute to the field of drug discovery. 

We will collaborate with the Functional Protein Research Group Bangladesh (FPRG,BD) of Dr. Sheikh Mahatabuddin, Associate Professor, Department of Nutrition and Food Engineering, Daffodil International University. They have the expertise to prepare peptide based drugs, protein engineering and chemically synthesized drug molecules and analyzing their effects in vivo and in vitro

The business model we would adopt is “Venture Capitalism”. In this system, we would start off the business with very little funding and a prototype of the platform would be built. We would then present this prototype to various investors who would be willing to invest a significant amount of their capital onto our system. As founders, we would retain complete organizational and operational control over how the business is conducted, but the investors would hold the larger share within the business. The investor-founder agreement would dictate to which degree the revenue would be shared between the two parties. However, it is imperative that operational control is retained by the founders, who have the subject matter expertise. We would actively and perpetually seek funding on a phased basis from various investors, thus providing our venture capital with a much more sophisticated business model. After certain stages of funding, the generation of revenue would switch from investor-based to a recurring model. 

Our vision about financial sustainability is a recurring revenue model. This is essentially a subscription-based system where the users of the platform would pay an affordable and nominal fee in order to use the course modules. Similar business models can be found on other course-based platforms such as Udemy and Skillshare. This is tantamount to our efforts to becoming the go-to platform for training and understanding these complex and modern concepts for pharmaceutical professionals within Bangladesh. It is to be understood that similar Monthly Recurring Revenue (MRR) based models have been used by various other platforms such as Netflix. Key concepts to understand in an MRR model: 

 1. Product-Market Fit: As there are no other organizations that are offering this service, we would be the first to enter the market, thus making us the perfect fit for this market.

 2. Recurring revenue: The business would have a recurring and incremental revenue system to ensure a constant flow of cash to the business. 

 3. Churn: The biggest nightmare in a Subscription based system. This occurs when a recurring and paying client stops paying and leaves the platform. Adequate measures must be taken to ensure churn is reduced. This may include hiring a dedicated account management team, who will consistently monitor high priority accounts to ensure that churn does not take place. 

 4. Tiered Subscriptions: Different levels of subscriptions (feature based) are to be modelled in order to sell the system. 

We are applying to Tiger Challenge to help understand a viable business model and secure funding, office space and technical support to establish a research institute. Tiger IT has already collaborated with MIT to ensure that a hackathon of innovative tools and problems are solved within Bangladesh. This is signatory of a trend towards the growth of technology-based solutions in the country. This puts us in the right market at the right time and provides us with just the opportunity to present our idea to the right market. The congregation would hold some of the technologically aware and entrepreneurial minds of the country and we believe that there is no better place than this to present our idea. 

We would like to partner with the following organization for tech and skill support. 

1. SCHRODINGER, a computational chemistry company based in New York, which includes the industry-standard PyMOL used for molecular visualization in their product catalogue. More importantly, they provide HPC and specialized software development services which will help us develop our own infrastructure.
2. CytoReason, a bioinformatics company specializing in machine-learning driven technology which models our cells and the immune system at a cellular level. Their work supports drug viability studies under experimental conditions, with models based on over a decade of research.
3. Insilico Medicine