As described above, PASS predicts simultaneously 3678 kinds of activity with mean accuracy of prediction about 95% (leave-one-out cross validation) on the basis of the compound's structural formula. So that you may use PASS for the prediction of the biological activity spectrum for existing compounds and compounds, which are only planned to synthesize.

Taking into account that the calculation of biological activity spectra for 1000 compounds in ordinary PC Core2Due 2,4 GHz takes about 10 seconds, one can effectively use PASS for predicting activity spectra of many compounds from large in-house and commercial databases.

Revealing New Effects and Mechanisms of Action

is considered below on the example of predicting the biological activity spectrum for well-known cerebrotonic drug Cavinton (Vinpocetin) launched by Gedeon Richter (Hungary) more than twenty years ago. Its structural formula and predicted biological activity spectrum are given below.

PASS application is useful because it gives the hits in the following:

Predicted biological activity spectrum for Cavinton.

45 Descriptors, 0 New Descriptors, 47 Predicted Activities

No Pa Pi Activity Experiment
1 0.929 0.004 Peripheral vasodilator  
2 0.900 0.000 Multiple sclerosis treatment  
3 0.855 0.005 Vasodilator +
4 0.844 0.003 Abortion inducer +
5 0.812 0.001 Antineoplastic enhancer  
6 0.760 0.006 Coronary vasodilator +
7 0.732 0.007 Spasmogenic  
8 0.700 0.036 Antihypoxic +
9 0.650 0.004 Lipid peroxidase inhibitor +
10 0.648 0.008 Cognition disorders treatment +
11 0.656 0.021 Antiischemic +
12 0.577 0.013 Acute neurologic disorders treatment +
13 0.540 0.039 Spasmolytic +
14 0.519 0.026 Antianginal agent  
15 0.486 0.037 Antihypertensive +
16 0.449 0.035 Antiarrhythmic +
17 0.432 0.063 Sympatholytic  
18 0.438 0.077 Sedative +
19 0.500 0.152 Antiinflammatory, pancreatic  
20 0.328 0.020 Antidepressant, Imipramin-like  
21 0.300 0.010 Thrombolytic +
22 0.342 0.075 Psychotropic +
23 0.276 0.023 Alpha 2 adrenoreceptor antagonist +
24 0.273 0.029 Anesthetic intravenous  
25 0.547 0.304 Vascular (periferal) disease treatment  
26 0.225 0.006 Antineoplastic alkaloid  
27 0.291 0.086 Cholinergic antagonist  
28 0.263 0.066 Benzodiazepine agonist partial  
29 0.417 0.238 Insulin promoter  
30 0.222 0.045 MAO-A inhibitor  
31 0.353 0.188 Cardiovascular analeptic  
32 0.249 0.100 Narcotic antagonist  
33 0.300 0.161 Acetylcholine release stimulant  
34 0.236 0.104 Antitumor-cytostatic  
35 0.271 0.165 Antiparkinsonian, rigidity relieving  
36 0.218 0.127 Antidepressant  
37 0.247 0.157 Analeptic  
38 0.211 0.126 Potassium channel antagonist  
39 0.243 0.158 Antiparkinsonian, tremor relieving  
40 0.333 0.258 5 Hydroxytryptamine 3 agonist  
41 0.233 0.172 Respiratory analeptic  
42 0.242 0.184 Antipsoriatic  
43 0.131 0.081 Analgesic, opioid  
44 0.147 0.128 N-cholinergic agonist  
45 0.285 0.267 cAMP phosphodiesterase inhibitor +
46 0.175 0.162 Anestetic general  
47 0.375 0.370 Male reproductive disfunction treatment  

As the Cavinton is used in medicinal practice for twenty years, many activities that were found in pre-clinical testing and clinical trials during this period are compared with the result of prediction. According to the available literature only 16 of 47 predicted activities of Cavinton are already found. These activities are marked by "+" in the Table above.

In particular, computer system PASS predicts the vasodilator and spasmolytic activities (Pa=0.855 and 0.540 respectively). It corresponds to well-known pharmacological effects of Cavinton, which causes the vasodilatation, increases the brain blood flow and metabolism. Antihypoxic and Antiischemic actions are also predicted for Cavinton (Pa=0.700 and 0.656 respectively). Really, Cavinton is used for these purposes. Cavinton is predicted as Lipid peroxidase inhibitor (Pa=0.650), Agent for cognition disorders treatment (0.648), Agent for acute neurological disorders treatment (0.577), etc. Cavinton has all these activities.

In predicted biological activity spectrum of Cavinton there are several actions which might become the basis for a new application of the substance. Among them: Multiple sclerosis treatment (Pa=0.900); Antineoplastic enhancer (0.812), Antineoplastic alkaloid (0.225) and Antitumor-cytostatic (0.236); Antiparkinsonian rigidity-relieving (0.271) and Antiparkinsonian tremor-relieving (0.243); etc. While the Multiple sclerosis treatment is predicted with high probability, all other additional activities have a relatively small Pa value. Thus, if these actions will be confirmed in the experiment, it might be the discovery of New Chemical Entities (NCE).

Similarly, the predicted activity spectrum for any compound provides the basis for its further testing. As a result some new effects and mechanisms can be found for old substances. Varying the threshold value of Pa one may choose the desirable level of novelty vs. acceptable risk of negative result.

Finding the Most Probable New Leads with Required Activity Spectra.

If a researcher can define which activities are desirable and which are not desirable for a compound according to the List of Activities (Supplement 1) predicted by PASS, she can select such compounds from the set of structures, which are available from in-house and commercial databases. For example, among the 15630 compounds from database of samples available in stock of ChemStar for which PASS prediction was carried out, 959 compounds are predicted as Endothelin antagonist, 236 compounds as Angiotensin II antagonist, 57 compounds as Angiotensin converting enzyme inhibitor. If the purpose of the study is to find the compounds with dual mechanism of Antihypertensive effect, e.g. Angiotensin converting enzyme inhibitor + Endothelin antagonist, only 11 compounds are predicted as having both activities. The best from the hits has Pa =0.170 (Endothelin antagonist) and Pa=0.244 (Angiotensin converting enzyme inhibitor). Based on this result one may decide either to test these 11 compounds or to carry out the prediction and selection for compounds from another database. In any case varying the cutoff value of Pa it is possible to choose the compounds with less or higher novelty (see: Interpretation of the Prediction's Results).

Selecting the Most Prospective Compounds for Highthroughput Screening.

If the required leads should have the activities which are included into the list of activities, predicted by PASS C&T, probably, the strategy considered in previous section is the best. However, such strategy cannot be applied in two situations: either the pharmacological target for which leads are searched is rather new and there are no compounds in the PASS training set related to this activity, or the Company would not like to disclose its fields of interests. In such case two other strategies are suitable.

The first strategy is based on the suggestion that the more kinds of activity are predicted as probable for a compound, the more probable to find any useful pharmacological action in it. For each compound from available set of samples the following value can be calculated:

where n is the number of biological activities under consideration (in PASS C&T n = 400). All compounds are arranged in the descending order of P values and compounds with the highest values of P which have the highest biological "potential" are only selected for screening.

The second strategy is based on the suggestion that the more is "novelty" of compounds relating to the compounds from the training set of PASS, the higher is probability to find NCE. Thus, the compounds with the highest amount of new descriptors have to be included into this sub-set.

Both strategies were tested on the datasets included about 100000 compounds totally and their efficacy is shown.

Determining the Screens that are More Relevant for a Particular Compound.

Based on the predicted activity spectrum for new compound, its testing can be organized in descending order of difference (Pa-Pi) for different activities. For example, if we consider the given above example of Cavinton, it should be studied in the following tests: Peripheral vasodilator (0.929-0.004), Multiple sclerosis treatment (0.900-0.000), Vasodilator (0.855-0.005), Abortion inducer (0.844-0.003), Antineoplastic enhancer (0.812-0.001), Coronary vasodilator (0.760-0.006), etc.

In this case both safety and efficacy of new compound will be characterized in more comprehensive way. Moreover, it is shown that the economic viability of such approach is more than 500%. Certainly, in a particular case a researcher will take into account also available facility of testing.