TACKLING NTDS AT COLLABORATIONS ...media.accelrys.com/.../collaborations-pharma-ntd.pdfthe majority of NTD lead identification has focused on phenotypic assays and, while many thousands
Post on 11-Jan-2020
0 Views
Preview:
Transcript
STREAMLINING DRUG REPURPOSING WITH BIOVIA DISCOVERY STUDIO
TACKLING NTDS AT COLLABORATIONS PHARMACEUTICALS
Customer Story
THE CUSTOMER: A LEADER IN “IN SILICO FIRST” SCIENCE According to the CDC, Neglected Tropical Diseases (NTDs) are a collection of parasitic and bacterial diseases that cause substantial illness (and many deaths) impacting over one billion people worldwide. These diseases dispropor-tionally impact the developing world, creating significant public health challenges for some of the world’s poorest and most marginalized people. Founded in 2015, Collaborations Pharmaceuticals target NTD and rare diseases. Their team of dedicated chemists, biologists and computational scientists have adopted an “in silico first” approach, applying a variety of machine learning techniques to identify novel and repurposed therapeutics. To date, they have received three orphan drug designations from the FDA and are constantly striving to provide safe and efficacious treatments for those who need them most.
THE CHALLENGE: STREAMLINING LEAD IDENTIFICATION The economic and scientific challenges of developing therapeu-tics have placed significant hurdles in the way of developing effective treatments for NTDs despite the acute need for them. In many instances, health organizations rely on preventative measures to slow the spread of disease. While this has helped to a degree, for many NTDs there is frequently little that can be done once infection starts. Outbreaks such as the 2014 and 2019 Ebola pandemics across Africa highlight the need for an increased pipeline of therapeutic candidates. However, the diversity of the pathogens implicated in NTDs, the relative complexity of their life cycles and their poor tractability in the lab have resulted in a dearth of potential therapeutic targets and candidates. Tackling NTDs, therefore, requires an alternative approach.
In recent years the pharmaceutical industry has gradually shifted towards target-based high throughput screening to identify lead candidates due to their cost and scalability advantages over whole-cell, phenotypic assays. For many NTDs, however, target-based screens would likely be ineffective, as poor characterization of their metabolic pathways and annotations of their genomes often produce few, if any, active therapeutic targets. As a result, the majority of NTD lead identification has focused on phenotypic assays and, while many thousands of compounds have been tested for a couple of these diseases – in this case for Ebola and Chagas disease – this has only explored a small fraction of the potential therapeutic space for these and many other NTDs.
THE SOLUTION: MACHINE LEARNING SUPPORTED BY BIOVIA DISCOVERY STUDIOCollaboration Pharmaceuticals’ approach to tackling this prob-lem focused on repurposing existing high throughput screening (HTS) data for Ebola and Chagas Disease to create machine learning models which could screen libraries for potential active
compounds, especially compounds previously approved by the US FDA for other diseases. BIOVIA Discovery Studio modeling and simulation software also supports chemistry-focused machine learning, providing an environment for the training and testing of a variety of methods, including Bayesian, Genetic Function Approximation, RP Forest and RP Single Tree. Further models developed in 3rd party applications such as R, in this case Sup-port Vector Machines, were also integrated into this workflow via BIOVIA Pipeline Pilot. These models sought to identify functional groups that correlated with highly active and inactive compounds in the HTS data. They utilized a variety of chemical descriptors such as molecular function class fingerprints of maximum diame-ter 6 (FCFP_6), AlogP and molecular fractional polar surface area. Discovery Studio allowed the Collaborations Pharmaceuticals team to assess the specificity and stability of these models with techniques such as 5-fold cross validation, ensuring greater trust in their results. Once trained models were performing acceptably, the team then could apply the models to screen new compound libraries. The leading compounds suggested by the model for each disease were compared to published pharmacophores in Discovery Studio, maximizing the team’s confidence that those candidates moving on to physical testing in various in vitro and in vivo studies would be the most likely to succeed.
THE RESULT: A REPRODUCIBLE METHOD FOR IN SILICO LEAD IDENTIFICATION As a result of the models created in Discovery Studio, the Collabo-rations Pharmaceuticals team was able to identify a collection of molecules that were progressed to in vitro and in vivo mouse stud-ies. These novel candidates, sourced from libraries of compounds approved for use by the US FDA, all possessed sub-micromolar EC50 levels of potency for both Ebola and Chagas disease. These results, validated both statistically and against known actives for both diseases, suggest that the models can effectively pre-dict potential candidate therapeutics in silico. Following these
Challenge:Identifying novel and repurposed therapeutic candidates for NTDs faster and more cost-effectively
Solution:BIOVIA Discovery Studio for chemistry-focused machine learning
Results:• Developed reproducible method for identifying
novel therapeutic candidates for NTDs• Produced <1µM actives for both Ebola and
Chagas disease with both in vitro and in vivo studies
• Achieved orphan drug designation for tilorone and pyronaridine as potential anti-Ebola therapeutics – also submitted for quinacrine for Ebola and pyronaridine for Chagas.
• Currently maintain a drug pipeline for 15 disease areas with a core team of 7 individuals
“Discovery Studio provides the technical depth needed for chemistry-focused machine learning. It has laid the foundation for a scalable solution to identify novel candidates for NTDs and a host of other diseases. This in turn has enabled us to obtain grant funding from the NIH to support these projects with our academic collaborators.”
— Sean Ekins, PhD, DSc, Chief Executive Officer, Collaborations Pharma
©20
19 D
assa
ult S
ystè
mes
. All
righ
ts re
serv
ed. 3
DEX
PER
IEN
CE®
, the
Com
pass
icon
, the
3D
S lo
go, C
ATI
A, S
OLI
DW
OR
KS, E
NO
VIA
, DEL
MIA
, SIM
ULI
A, G
EOVI
A, E
XALE
AD
, 3D
VIA
, B
IOVI
A, N
ETVI
BES
, IFW
E an
d 3D
EXCI
TE a
re c
omm
erci
al tr
adem
arks
or r
egis
tere
d tr
adem
arks
of
Das
saul
t Sys
tèm
es, a
Fre
nch
“soc
iété
eur
opée
nne”
(Ver
saill
es C
omm
erci
al R
egis
ter #
B 3
22 3
06 4
40),
or it
s su
bsid
iari
es in
the
Uni
ted
Stat
es a
nd/o
r oth
er c
ount
ries
. All
othe
r tra
dem
arks
are
ow
ned
by th
eir r
espe
ctiv
e ow
ners
. Use
of a
ny D
assa
ult S
ystè
mes
or i
ts s
ubsi
diar
ies
trad
emar
ks is
sub
ject
to th
eir e
xpre
ss w
ritt
en a
ppro
val.
Europe/Middle East/AfricaDassault Systèmes10, rue Marcel DassaultCS 4050178946 Vélizy-Villacoublay CedexFrance
AmericasDassault Systèmes175 Wyman StreetWaltham, Massachusetts02451-1223USA
Asia-PacificDassault Systèmes K.K.ThinkPark Tower2-1-1 Osaki, Shinagawa-ku,Tokyo 141-6020Japan
Our 3DEXPERIENCE® Platform powers our brand applications, serving 11 industries, and provides a rich portfolio of industry solution experiences. Dassault Systèmes, the 3DEXPERIENCE® Company, provides business and people with virtual universes to imagine sustainable innovations. Its world-leading solutions transform the way products are designed, produced, and supported. Dassault Systèmes’ collaborative solutions foster social innovation, expanding possibilities for the virtual world to improve the real world. The group brings value to over 210,000 customers of all sizes in all industries in more than 140 countries. For more information, visit www.3ds.com.
findings, the Collaborations Pharma team submitted and received approval for orphan drug designations for two of the candidates, which were discovered with the Ebola model, tilorone and pyro-naridine. This designation can help accelerate the development and commercialization of these compounds, offering much needed aid to those who suffer from this disease and potentially minimizing the risk of future outbreaks. At the time of writing, further submissions for quinacrine as a treatment for Ebola and pyronaridine for Chagas disease are under review. As a result, the Collaborations Pharmaceuticals team has developed a compel-ling and reproducible method for the identification of new drug compounds for NTDs and other diseases in silico. To date they maintain a drug pipeline for 15 disease areas with a core team of 7 individuals, demonstrating the power of their “in silico first” approach to drug design.
ACKNOWLEDGEMENTSCollaborations Pharmaceuticals would like to thank the National Institutes of Health for funding support on these projects through grants NCATS R21TR001718 and 1UH2TR002084-01.
DS-9895-0819
REFERENCES1. https://www.cdc.gov/globalhealth/ntd/index.html.
Accessed 7/22/2019.
2. Ekins S, Freundlich JS, Clark AM, Anantpadma M, Davey RA, and Madrid P. Machine learning models identify molecules active against the Ebola virus in vitro [version 3; referees: 2 approved]. F1000 Research, 4:1901. 2017.
3. Ekins S, Lage de Siqueira-Neto J, McCall LI, Sarker M, Yadav M, Ponder EL, Kallel EA, Kellar D, Chen S, Arkin M, Bunin BA, McKerrow JH, and Talcott C. Machine Learning Models and Pathway Genome Data Base for Trypanosoma cruzi Drug Discovery. PLOS Negl Trop Dis, 9(6): e0003878. 2015.
4. Lane TR, Comer JE, Freiberg AN, Madrid PB, and Ekins S. Repurposing Quinacrine Against Ebola Virus Infection In vivo. Antimicrob Agents Chemother. 2019 Jul 15. pii: AAC.01142-19. doi: 10.1128/AAC.01142-19.
top related