Leishmaniasis

Leishmaniasis is a neglected disease. As a disease of poverty, it has historically received limited funding for discovery, development and delivery of new tools. Current treatment is costly, lengthy, painful and sometimes toxic. Like a handful of similar diseases, it is the scourge of whole regions affected by them, because we still miss cheap, safe and effective cures for them.

At the same time, new drug candidates are being developed and old ones are being tested every day. Today, millions of drug activity assays are available at the press of a button. In this Indaba Grand Challenge, we dare to ask you to help identify amongst the already known, tested and (often) approved drugs, potential cures for different forms of leishmaniasis.

Drug Repurposing

Drug repurposing is the act of researching if an existing drug can be used for new therapeutic purposes. It’s something that has been done for centuries. A large amount of medications on the market right now are used for multiple purposes: for instance, aspirin is a pain reliever and also a blood thinner, ibuprofen is also used for pain relief, and for reducing inflammations.

The high cost of research and development involved in finding new drugs means that most of these will be targeted to “First World” countries to generate returns.

For “neglected diseases” or “tropical diseases”, drug repurposing is used as a sort of universal strategy: it lowers the costs of research and production and the time needed for clinic trials, which is why you’ll find that many of these diseases use medication that was originally used for other illnesses.

THE CHALLENGE

Drug discovery methods and pipelines are well described and modular. Each can be decomposed into smaller pieces, which can be rearranged and modified. All of these pipelines rely heavily on knowledge representation, pattern detection and building of statistical models.

The goal of this Indaba Grand Challenge is to, through the use of standard techniques and publicly available data on drugs, drug targets and kinetoplastids (parasites causing amongst other leishmaniasis), identify either:

• Compounds among known, approved drugs which can target known drug targets in Leishmania (especially one of the 15 species known to cause the cutaneous type of disease) in addition to their official use.
• Compounds among known, approved drugs with sufficiently similar targets in the goal organisms to the targets in kinetoplastids.

Importantly, the verification of a drug-target interaction is possible and routinely done in silico. However, not all drug-target combinations can be verified, due to sheer combinatorial expansion and limited computational resources. In addition, the identification of similarities and analogies between known experimental data is far from trivial and, despite the existence of numerous approaches, cannot be solved by one statistical technique alone.

By launching this Grand Challenge, and true to the core values of Deep Learning Indaba, we aim to build a collaborative effort, joining experts and enthusiasts in machine learning and software engineering, with doctors, epidemiologists, medicinal chemists and policy experts, to create a synergistic effect at scale. Our vision, if it becomes true, is one where interesting discoveries of the community are taken up by the experts, and conclusions of the experts are fed back to the community to iteratively improve on its findings and make breakthroughs possible. No single person can do it perfectly, but through a joint effort, together we can.

CHALLENGE SPECS

Goal: Propose a new treatment, comprising an Leishmania protein (either from a defined species, or a protein present in the proteome of one or more of the Leishmania species) and a small molecule (or set of small molecules). Submission can be specified as:

• Small molecule (SMILES/SMARTS string or SDF file) and a protein name
• PDB file with a structure of a target and a bound small molecule

Available resources: We have put together a list of available resources and additional material here.

WHO CAN APPLY?

We invite you to join this initiative and help put an end to Leishmaniasis. We need your involvement, especially if, you are:

• Data science or Machine Learning expert or enthusiast: we have the data, tools, benchmarks and everything you need to know to attack this problem. Our community needs your data wrangling skills. If you feel up for a challenge, apply here!
• Bioinformatics and Cheminformatics people: in addition to the resources, data and standardized protocols, we have a community, which can help you go the final mile from a great protocol to a great candidate. We need your expertise in translating life into bytes. If you can help, please apply here!
• Kinetoplastid researchers, practitioners and clinicians. We need you dearly. We will be generating loads of hypotheses and we need you to help make sense of them. We need you to answer questions, we cannot answer ourselves. In return, we will give you a fresh, new insight to the problem you care about. Please apply here to help us cure this scourge!
• Universities, institutes and research organizations. We know it is difficult to find either meaningful project work for students, or dedicated people to do the work with you. If you feel your students would enjoy working with cutting edge methodology, in a collaborative discovery environment, that includes everyone, please do not wait any longer with contacting us. If your research can contribute to this project, we would be deeply grateful, if you could join us. Even if you want to bounce some ideas, we want to hear from you. Apply here and join the truly collaborative effort to democratise drug discovery!
• NTD-oriented organizations and policy makers: We need you more than ever – this project will result in viable drug candidates. We will need a lot of good will to get them to the people in need. We also need access to those, for whom African trichomoniasis is their life work, experts and field workers. We need their expertise. You have the ear of those who matter for us the most. We work to help advance your cause. Together, we can make a difference. If you believe that together we can make a difference, please do apply here!
• Companies in Machine Learning/AI, Bioinformatics or other: If you’re interested in making a valuable contribution to the success of this Grand Challenge, there are multiple ways you can help. What’s straightforward but important is to provide cash or cloud computing credits, as this is a quantitatively intensive project that will make extensive use of GPUs and cloud compute power. You can also contribute in kind with access to your platform or technical expertise like some of the companies already participating in this initiative (see list below, or even better join the list).

If you are excited about this project, but do not belong to any of the categories listed, if you have comments, questions or just want to say hi – we want to hear from you! There is no voice too quiet or contribution too small to make a difference. We need community ambassadors, spokespeople, word-of-mouth champions. We need everyone. We need you, let’s make it happen!

Here is more about the challenge and how to register