1. What is unique and special about your platform?
Our methods are grounded in the fact that, in nature, there exist only about 500 unique binding pocket shapes. We copy these pockets, as necessary, to predict the binding properties of proteins which lack crystal structures. This approach is many times more computationally efficient than traditional methods which rely on the existence of experimental high resolution structures or close homology models, which are only available for 25% of the human proteome. PanXome effectively predicts binding to 97% of the human proteome, enabling us to expand the boundaries of quality hit predictions and also predict off-target interactions with a high degree of precision. This ability to predict side-effects and safety is an unprecedented capability in the industry.
2. What is the precision of your virtual ligand screening (VLS) and how does it compare with competing models?
We have experimentally validated the precision of our VLS predictions and found that, on average, our precision is about 30%, when screening our top 100 predictions. This is about 10-20 times better than competing structural-based approaches.
3. What is the precision of side effect predictions and how does it compare with competing models?
We predict whether or not a drug will pass a Phase 1 safety clinical trial with 80% precision (at least one catastrophic side effect predicted). We predict all side effects with 60% precision.
4. I have a small molecule. What do you need from me and what do I receive back?
We just need the chemical structure of the small molecule; preferably as a Smiles string . We deliver back to you the name and models of all proteins that the small molecule interacts with, a list of possible side effects, the Kill Index, the number of proteins with severe side effects that the small molecule is likely to interact with and predicted binding modes to the top protein targets.
5. I have a protein target. What do you need from me and what do I receive back?
We just need the name of the protein and the protein's sequence and if available, the protein's structure. You can provide a proprietary small molecule library. If not, ZINC15 and DrugBank molecules will be screened. We deliver a model of the protein, a list of possible small hits, a list of off target proteins/hits, a list of possible side effects for the top scoring small molecules and the Kill Index (the number of proteins with severe side effects the small molecules are likely to interact with).
6. I have a disease target. What do you need from me and what do I receive back?
If no protein target is known, we can predict a set of proteins associated with the indication of interest. We will then coordinate with you to select the most promising protein targets. You can provide a proprietary small molecule library as well. If not, ZINC15 and DrugBank molecules will be screened. You will get a model of the protein, a list of possible small hits, a list of off target proteins/hits, a list of possible side effects for the top scoring small molecules and the Kill Index (the number of proteins with severe side effects the small molecules are likely to interact with).
7. Does your platform work on small molecules and biologics?
The platform is optimized for small molecule virtual screening and side-effect predictions. In general, the current technology is not applicable to biologics although we can assist in analysis of protein-protein interactions in specialized cases.
8. What are the commercial terms of your service?
We offer commercial terms depending on the needs of the client:
A) Fee-for- Service – An analysis report is provided for a fixed fee.
B) Software as a Service (SaaS) – For high volume work or when the client does not wish to disclose its proprietary molecules or targets with PanXome, PanXome can make its tools available in a cloud environment such as Amazon Web Services (AWS).
C) Risk-shared Collaboration – This model is typically reserved for engagements where PanXome works collaboratively with the client to advance and co-invent new intellectual property. The compensation may take the form of some combination of a technology access fee, fee for services, success-based milestones and downstream royalty.
Please contact us so that we can tailor a commercial offering to suit your specific needs.
9. How long does it take to obtain my results back from PanXome?
For specific projects involving a small number of targets or molecules, we typically provide the final report in two to four weeks.
10. How do I initiate the process?
Please send us a message through the Contact Us section of the website and a PanXome representative will contact you.