|
|
This is how a drug against a worldwide epidemic is developed |
There are many diseases that threaten humanity, and only recently we witnessed
an outbreak of avian influenza. Matrix Pharma's chief scientist talks
about what goes on behind the scenes in the innovative pharmaceutical industry
Dr. Marcel Thuerk
The fear of a worldwide outbreak of avian influenza is keeping the
health authorities and many scientists in the world awake at night. Although
bird flu is no longer the opening item on the news, the danger it
poses has not passed. On the contrary: if, before, they had feared the
moment when the virus causing the disease would mutate and infect
people – that moment has already arrived.
The death rate among people who are infected with avian flu is particularly high at 50%.
To date, according to reports from the World Health Organization, approximately 154 persons
have died of the disease, most of them in Asia, and there are possibly many more
cases that were not reported. Recently the virus progressed to an even more dangerous
stage: it has mutated and can now be transmitted from one person to another. This is
what happened this summer to an entire family in Indonesia, whose members infected
each other.
In terms of the significance of a worldwide epidemic, the estimates are most alarming.
Various governments and the World Health Organization publish rather different estimates,
but even the most optimistic of these notes the anticipated death of millions of
people all over the world and costs to the world economy running into the trillions.
I don't know when it will happen, but what is clear is that a solution is vital, and quickly.
At the time, countries prepared large stocks of drugs against the virus, but these will be of
no good when the epidemic breaks out. The drugs are no longer effective against the
current mutation of the virus, which has already undergone many changes.
A long, protracted process
The process of developing a drug is very long, and it takes several years until the approvals
are received from the authorities and the drug is marketed. In actual fact, less than
10% of drugs reaching the clinical trial phase are indeed approved by the authorities,
mainly as a result of the medication's toxicity or partial efficacy. The computerized processing
method applied by our software allows a large number of parameters to be tested
simultaneously. The platform is based on artificial intelligence, which simultaneously optimizes
multiple parameters in all phases of the development and screening of the source
materials tested in the laboratory, until those substances with the best chances of success
in the pre-clinical and clinical phases are identified.
When you are developing a drug, the question is whether you need the knowledge and
expertise accumulated during hundreds of years, or, if you are stuck with something new,
which you don't know much about – can you start to research it anew without the earlier
knowledge? We have developed a software-guided laboratory process called CADDIS –
computer-aided drug discovery – that is capable of creating the knowledge necessary for
drug development, for developing the tests and examining their results. In fact, the software
does not use prior knowledge, the laws of physics or chemistry, but is given the test
data, the descriptions of the molecules, and it analyzes the structure of the correlations
between them. This process leads to the development of multiple drug "leads" (compounds
that have already proven themselves in preliminary laboratory experiments), efficiently
and methodically.
This is how it works: lab results from research related to the bird flu virus are fed back
into the computer system, which is able to measure the required effect of a variety of
compounds. The software learns and understands the information fed into the system and
is able to explain the chemical common denominator of all the compounds and in this way
provides different proposals based on the explanations derived from the information and
from data stored in the information repository of the compound directory. These compounds
are synthesized and returned for use in the lab in another round of tests. This
process is repeated six to eight times (cycles), and at its end completely new compounds
are created, ready to go into pre-clinical and clinical trials.
The lab tests were designed in a manner that all the drugs are optimized simultaneously,
and the compound that is chosen is the one with the greatest action, the highest selectivity
(i.e. less side effects), capable of being swallowed, possessing metabolic stability (i.e.
the body does not alter its chemistry), good serum protein (so that it is carried to all parts
of the body), low toxicity, and any and all other necessary criteria defined as important for
a particular therapeutic indication.
Getting out of the box
The key to finding a solution to bird flu also lies in a change of approach: instead of trying
to defeat the virus, which requires knowledge of each and every mutation, we adopted a
different line of thought. In the project of discovering the bird flu drug our approach is not
to inhibit a particular enzyme of the virus, similar to treatment with neuromidase inhibitors
such as Tamiflu, but rather to act on an unspecific mechanism in the host cells and
eliminate the virus's ability to multiply.
We work with a lab that has access to the range of viruses and in a secured facility that
enables the job to be done. We have already begun the process of screening the effects of
the first round of random virus compounds inseminated in ova. Different viruses are implanted
in the ova, and the compounds are subjected to revaluation on the basis of their
ability to equally block all viruses implanted. Meaning, the idea is to develop a drug that
will affect the host cell, and not the virus. Each ovum is also treated with the drug to
make sure that the compound has no toxic effect.
The project has three phases: in the first phase, ten compounds of known drugs are subjected
to a screening process for efficacy, and of them, only a number of anti-viral compounds
are selected. The compounds are then tested for a second time against three influenza
strains and two RNA virus strains. In the second phase, 100 small organic molecules
and 100 small peptides chemistry are tested four times on sensing virus strains, using
control indices. In actual fact, the number of laboratory tests for 200 compounds will
reach 5,000.
In the third phase, the results are fed back into the CADDIS software, which suggests 25 small organic molecules and 25 peptides for retesting. In this phase too, the compounds Extracting Knowledge out of Data are tested for toxicity, biological availability, protein binding, etc. This cycle is repeated six to eight times with fifty enhanced compounds, until satisfactory effective lead compounds are obtained.
Dr. Marcel Thuerk is chief scientist at Matrix Pharma, a drug development
company applying an innovative, artificial intelligence based method
Published in "ynet science",
January 3, 2007
|
|
