Hello everyone, this is a course on Computer aided drug design.

I am going to take 40 lectures each lecture of half an hour duration.

So we are going to talk about how to design drugs with the aid of computers, so computers

will be an aiding tool, but you cannot replace a medicinal chemist or a synthetic organic

chemist or a physical chemist in helping in designing new drugs.

So this particular talk is going to introduce this topic of Computer aided drug design.

So what is this drug discovery, so this is the process by which a drug candidate is identified,

it is partially validated for the treatment of a specific disease.

We are going to look at why partially validated because unless you actually do an animal trials,

unless you actually do human volunteer trials validated compound does not become a drug

which can be introduced in the market.

So that is why we call it partially validated.

So you need to know the mechanism of action, you need to know how the drug acts, what are

the enzymes on which it goes and binds to, or a protein there by inactivating the particular

pathway of the disease, so we need to understand the mechanism of action.

So we need to identify that particular target, the enzyme or protein, and we need to validate

the target that means we need to use tools to be sure that that is the target on which

my drug is acting.

We need to identify a candidate.

Generally before the compound is approved by FDA and introduced in the market, it is

not called the drug it is called a lead compound that means a lead, a compound which has shown

very good activity, a compound which looks very, promising, so that is called the lead

identification okay.

And then we need to optimize the lead because just because the compound acts very well in

my lab does not mean it has all the properties of a drug.

As you know drug is something that is taken inside the human body, so it should not be

toxic, it should not cause side effects, it should have a good absorption in the stomach

and so on.

So they are called drug likeness property.

So we may have to modify those properties, so that it not only has good activity but

also has good drug likeness property.

So we are going to introduce this term drug likeness property for a many, many lecture

okay.

So it should have these drug likeness property and it is called ADME, that means A refers

to absorption, D refers to distribution, M refers to metabolism and E refers to excretion.

Let me write down absorption, distribution, metabolism and excretion.

That means the drug has to get absorbed into my stomach okay, mostly if it is an oral drug

it has to be taken it goes to GI track that means gastrointestinal tract gets absorbed.

Then it gets distributed inside your bloodstream, it gets metabolized because there are many

enzymes your liver is a gatekeeper which tries to keep on degrading any foreign compounds

coming into it, and then finally it has to get excreted that makes thrown out of your

body.

So the drug which is taken in has to have good ADME properties, that means absorption,

distribution, metabolism and excretion properties.

So it may be a very good active compound but maybe gets metabolized and becomes inactive

inside your stomach.

So then it is of no use okay, So ADME is also very important when you are designing drugs.

Then it should have good PK and PD, what is this PK and PD?

PK is nothing but pharmacokinetics and PD is nothing but pharmacodynamics.

That means how the drug is excreted as a function of time?

How the drug is absorbed as the function of time into the plasma?

That is all given by pharmacokinetics and how the drug which is inside the body acts

on the disease that is called pharmacodynamics.

We are going to talk about each one of them much more in detail, and it should not have

toxicity.

So we need to know the toxicity of the particular compound which we are testing that is the

lead which we are testing, it should be non-toxic short-term toxicity as well as long term toxicity.

For example, if somebody has an arthritis problem they may have to take the anti-arthritis

drugs for a very, very long time, so it should not have long term side effects or toxicity

okay.

So all these properties need to be looked at when we are designer drugs, it is not only

the activity in my lab, but I need to understand what is the mechanism of action, which target

it goes and binds to, and prevents the disease progression.

It should have good properties, drug likeness properties like absorption, distribution,

metabolism and excretion, it should have good pharmacokinetics and pharmacodynamics, it

should not have any toxicity and so on.

So a drug discovery process does not include many steps: preclinical studies, so I have

a nice lead in my lab, so I do not introduce that immediately into the market, they need

to be something called a preclinical studies.

Normally, that is done in with the help of animals maybe rats, rabbits, dogs or monkeys

and so on that is called preclinical studies.

Then comes clinical trials where we use a human volunteers, then comes the regulatory

approval you need to get approval from the regulating body like food and drug administration

authority, and then it goes into sales and marketing.

So these are entire drug development process actually.

But we are not going to talk about all these, our focus is more on these rather than these

okay.

So it is much more preclinical studies we are going to stop at.

So let us look at the history of computer aided drug design, it started in 1960s okay.

So everybody started looking at the target, how the drug which target it went and bound

to before those 60s yet drug was given to the patient, and there was a reduction in

the pain or reduction in the fever or reduction in the infection, and that is it.

Nobody bothered about how the drug reacted, which targets it went and found to okay.

Then came the 80s where they started using high throughput screening by which many, many

candidates where tested for certain targets okay in an automated way 10000-20000 compounds

were tested in one shot by pharmaceutical companies okay, that is in the 80s because

it is not possible to take one compound and keep testing one after another which may be

a very time consuming process.

Then in the 80s okay again databases started coming, lot of databases are created combinatorial

databases, database on activity of large number of compounds, activity of compounds on different

targets, they are called combinatorial libraries.

Then again fast computing started coming into use.

Computing power increase, parallel processes came, so scientists started using different

parallel processing machines for performing ligand protein docking, drug protein docking,

drug target docking.

Then again 90s genome assembly genomic based target selection, that means we try to understand

what are the genomes involved in the disease processes.

Then in the 2000 we went into pharmacogenomics, that means combining genomics with pharmacological,

so which genes affected, which pharmacological issues.

So this I would say approximately the history of a Computer aided drug discovery.

Now let us look at some names of companies which are involved in drug discovery, and

of course this is more as a overall knowledge or picture.

The top 10 pharmaceutical companies you might have heard about it, this is 2014 data: Novartis,

Pfizer, Roche, Sanofi, Merck, Johnson and Johnson, GlaxoSmithKline, AstraZeneca, Gilead,

Takeda.

And look at their annual sales in million dollars, this is almost 47 billion dollars

45 billion dollars.

So if you add up all these you are talking in terms of 200 to 300 or even 500 billion

dollars okay that is a big business.

So a drug discovery drug business is very, very large, these are some bio-pharmaceutical

companies.

Now many drugs are being manufactured using biological route rather than chemical route,

and so many bio-pharmaceutical companies have come selectively using biological routes.

Whereas now these companies which were making dugs through chemical routes have also started

making some products using biological routes, some of these companies are Amgen, Genentech,

Serono, Biogen, Genzyme and so on actually.

So a drug discovery if you look at it the global scenario, the market is very large.

US is the biggest market followed by Europe, Japan and so on.

Let us look at Indian companies they are much smaller when compared to, for example Novartis

or Pfizer maybe 100 size, but still they play a very important role, especially in generic,

Ranbaxy Dr Reddy's lab, Torrent, Cadila, Cadila Healthcare, Cadila Pharma, Lupin, Sun Pharma,

Nicholas Piramal, Dabur.

So all these are Indian companies, and as I said India plays a very important role in

generics that means drugs which have come out of the patent regime which anybody can

make.

Indian industries are extremely competent in making compounds in a much cheaper way,

so they become, the products become very competitive and they are able to export to US, Europe,

third world countries and so on actually.

So Indian companies are not into new drug discovery, but they are into the generics

and over the counter drugs, and they are really competing globally.

And their products are approved by FDA that means food and drug administration authority

and so on.

So let us look at the top 10 selling drugs year 2013, just gives you an idea of what

does drugs currently which are very, very important okay.

So you look that these the top selling drugs in 2013 is an anti-inflammatory drug for rheumatoid

arthritis.

14.8 billion sales okay that is a big money.

Then this is also here something to do with arthritis autoimmune manufactured by these

companies that is also very big.

Then comes another third compound which is also for rheumatoid arthritis.

Then we have a drug for asthma, diabetes, cancer, so several drugs for cancer, and then

cholesterol lowering drugs and so on.

So all these top selling drugs as of 2013, they are all in the range of 14 billion going

right up to 5.5 billion sales.

And the other drugs which are marked in red are manufactured using chemical routes, whereas

other drugs are manufactured by biological routes.

So what it means is slowly the pharma companies are moving towards making drugs using by a

bioprocess approach rather than chemical approach, when you say bioprocess we are using either

an animal cell or a plant cell or a bacteria or virus or yeast cell and so on to make the

product.

Only 3 compounds are fully manufactured using a chemical route, I think as time progress

all these molecules maybe manufactured using biological route.

So the lecture which I am going to give in the next 40 lectures will cover topics like

structure and property, so what are the structural features that a drug should have which will

lead to certain property.

For example, property could be solubility; property could be its lipophilicity and so

on actually.

Even properties like toxicity, degradation, stability at different pH condition, they

are all called properties.

And then drug likeness, does it have the drug likeness property, that means just because

I design a molecule and it shows very good activity does not mean it can be taken in

by human, because like I said it should be less toxic, it should be not have side effects,

it should be coming out of the body after it is done its function, it should be stable

at various pH conditions, so these are all called drug likeness property, so we are going

to talk about that.

Then of course absorption how is it getting absorbed into the body, distributed, metabolized

excreted and then oral bioavailability.

So we take the drug orally, so that means it goes into the stomach where the stomach

is highly acidic we are talking in terms of pH=2, and then it gets absorbed into the stomach,

it goes to the blood and then from the blood it gets circulated into the body, and the

liver which tries to degrade whatever a foreign objects there.

So it has to overcome that and finally has to reach the target site, so that is called

oral bioavailability.

So just because I take a 10 milligram tablet does not be at the target site the concentration

of the drug will be 10 milligrams, because it might not be absorbed properly in my stomach

or it may get degraded in the acidity in my stomach, or it may get degraded by the liver

enzymes.

And so it may whatever concentration that is reaching my target site could be very,

very small, so that is called oral bioavailability, so we are going to talk about that.

Then we come to force fields, what are the different force fields we use Well, molecular

modelling, and then how do we calculate minimum energy confirmation?

How do we minimize the structures of molecule?

What are the boundary conditions we use?

QSAR that, is quantitative structure

activity relationship.

Then we are going to look at the structural features of the drug that is called pharmacophore.

Then we are going to look at how to modify the structure that is called scaffold hopping,

and then we look at how targets are being identified, and how do we identify how the

drugs goes and binds to a particular target.

So we are going to look at docking, and then we are going to look at pharmacokinetics and

pharmacodynamics that PK and PD okay whatever I said.

So in the next 40 lectures, we are going to look at all these topics.

We are also going to look at certain softwares which are freely available, which we can use

for performing all these operations.

There are many, many softwares which are freely available, so we do not have to buy any commercial

software, I am not going to talk about any commercial software, but all the freely available

softwares we can use to perform several of these operations.

There are very good freely available softwares in the web, we are going to look at all of

them.

So what are the computational resources for this, there is something called SWISS ADME

I have given this site address also, so you people can have a look at it.

One is called SWISS ADME; when we draw a structure it can give properties of molecules; it can

give what is the oral bioavailability.

Then we have this SWISS target prediction.

It can predict what are the possible targets if I give a structure, that is also very useful

actually.

This one, if I give a molecular formula it gives you a molecular structure.

This software predicts the pharmacokinetics properties of a drug candidate using QSPR.

Then this software predicts toxicity of a molecule, and also toxicity of the fragments

of the molecule.

Because when a drug goes inside the body as I said the enzymes in the liver breaks the

drug into small fragments, so the fragments also could be toxic.This software tries to

predict whether the drug itself has toxicities, properties or the fragments that are created

generated by the degradation of the drug also has.

Then we have QSAR tool box.

Then we have a chemoinformatics, ADMET, physical chemical predictions, this this software called

VLS3D okay.

And then there is something called 3DQSAR, 3 dimensional QSAR, there is a software for

this.

Software gives you the pharmacophore, details that means it can identify the features of

the molecules and try to search millions of chemical structures okay, which have the same

pharmacophore.

This another software gives you the QSAR pharmacophore model.

Then we have a property prediction, OSIRIS.

Then we have a another property prediction that is called a eDragon, and there is another

software for property protection that is called Molinspirational.

So you see a lot of softwares available okay, and during the course I am also going to use

many of these softwares, and I am going to demo to you people how to use some of them

based on my experience and so on actually.

Then there is another software is called the therapeutic target.

So it has data information about the therapeutic protein nucleic acid targets described in

the literature including the targeted disease conditions, the pathway information and so

on.

Then there is another software called STITCH, so it connects the chemicals or the drug with

the target, and genes so it creates a very complicated map.

It stitches basically between the genes the target the drugs, and creates a very complicated

map connection.

So it can connect almost 1.5 million genes.

Then there is a T3DB toxin, toxin target database, so the common toxins and their associated

toxin targets we can identify using this database.

So lot of softwares, and I am sure you guys maybe finding your own software as you keep

searching the web.

Books, there are some good books are there; Drug-like properties: concept structure design

methods ADME to tox okay this is a very good book.

And then you have guidebook on molecular modelling in drug design, Drug design, Pharmacophore

perception, Silico drug design, Software and resources computation.

So very good books are there you guys can have a look at them.

Databases, we have this ZINC database.

ZINC has 35 million purchasable compound structures, so if I going to look at those structures

those properties and I find one molecules looks very promising I can even buy from them,

for a cost.

So this is a very useful database.

Then we have the drug bank database, there are 8200 entries in that, that means drugs

which are approved by the food and drug administration authority, that is FDA approved drugs.

Then Biotech drugs, nutraceuticals experimental drugs.

So if I want to know is there are any drugs for a particular disease I would look at this

particular database or another database called drugs.com.

Thise got 24000 prescription drugs over the counter medicines and natural products.

So first step if I want to know whether there is any drug available for a disease is that

I will go through these 2 databases.

Then you have Chemspider, ChemDB, chemoinformatics tool, PubChem, ChemMBL, all these databases

give structures of molecules, structures of drugs, structures of prescription drugs or

structures of nutraceuticals, and some properties like solubility and lipophilicity, molecular

weight, number of rotatable bonds, number of hydrogen bonds and so on actually.

So these are I would call databases very useful, for if you want to look at structures.

More of it pharmacogenomics knowledge implementation, this another database is called PharmGKB,

This resource that curates knowledge about the impact of genetic variation on drug response.

So if there is a genetic variation because of different races or different communities,

then how the drug will respond, so there is a lot of information, data on genes, diseases,

drugs, pathway.

So we can look at this database also if you are interested.

Then of course MedChem designer, so large number of databases structures of billions

of molecules are available as a resource.

More of this databases of interest for drug discovery okay.

So we have of course as you can see some of them are commercial products, some are publicly

available, freely accessible for academia, commercial versions also available.

So we can have large number of databases on molecules of different structures, of course

some of them are commercial also.

That means if you make a payment you will be able to get databases for you which can

be used for in silico analysis.

Journals, these are some few journals that are talking about Computer aided drug discovery.

Medicinal chemistry, Drug discovery today, some of them deal with reviews and some of

them deal with experimental studies and so on.

So large number of resources I have introduced it to you, you can play around look at some

of these structural databases as well when we have time , and see how useful they can

be okay.

Let us look at some structures, aspirin this is a molecule we all must be knowing for a

very long time, and we all would have taken this particular drug, it is an analgesic,

it is a pain reliever, it is an antipyretic that means it reduces fever, anti-inflammatory

that means it reduces inflammation and anticoagulant.

Aspirin has become a wonder drug, nowadays it is being used somebody has a problem with

cardiovascular problem.

And if the blood viscosity has to be reduced they are asked to take aspirin, so that is

also used as an anticoagulant.

So it inhibits the production of prostaglandins, these prostaglandins are mediators which carry

pain, this is the structure of aspirin is the acetylsalicylic acid okay, acetyl this

is the acetyl group okay, this is the salicylic acid.

There are many variations of aspirin because you can think about differentsubstitutions

and you can have different structural features.

So if you want to go into a drug discovery, we are talking in terms of 10 to 15 years

it cost about 1 billion dollars, you may have to test maybe 10,000 compounds, animal trials

or clinical trials or in the lab, and then you may have to test it out on about 1000

to 5000 volunteers, and hopefully you may have one successful candidate okay.

So drug discovery is like a lottery.

In computers we may have to test maybe 100 thousand compounds, in the lab we may have

to test on about 5000 to 10000.

And finally you may succeed with one drug, or some of them or most of them will fail.

So a lot of money is lost because of this particular exercise that is why when drugs

are introduced, a newly introduced drug is always very, very expensive, because there

is lot of money involved, and there are many failures for every drug introduced into the

market.

So in the 60s drug discovery used to be cheaper, talking in terms of 4 million US dollars,

but as you can see you can go down it has become almost a billion dollar.

Simultaneously, in 50s maybe only few compounds were tested, but as you go down we can see

we are talking in terms of 10 times more compounds are being tested.

So we started testing more compounds the cost of marketing and drug also keeps going up

and up and up.

And there are only 6000 known drugs, but there are 10 power 60 possible compounds.

So obviously there are many drugs in this compound list which we do not know what type

of activity it will have.

So there is lot of scope for developing a new molecules and new compounds and so on.

So we will continue in the next class of the concept of drug discovery, and what are the

steps involved in target identification and lead identification, thank you very much for

your time.