WO2002014538A2

WO2002014538A2 - Rational panel design method for microbiological investigations

Info

Publication number: WO2002014538A2
Application number: PCT/US2001/025729
Authority: WO
Inventors: Alan Mark Hochberg
Original assignee: Qualicon, Inc.
Priority date: 2000-08-17
Filing date: 2001-08-16
Publication date: 2002-02-21
Also published as: WO2002014538A3; AU2001283422A1

Abstract

The invention is an interactive, rational panel design system over a computer network without human intervention other than by the customer. A network-based, computer-implemented method for the design of screening panels of organisms takes advantage of the worldwide presence of culture collections and databases which are forged into a network providing genetic data to a database basic to the process. Based on criteria supplied by the customer, the associated computer at a location remote from the customer, designs the most economical panel to submit to the customer. In a subsequent operation the customer may order one or more of the selected organisms.

Description

TITLE

RATIONAL PANEL DESIGN METHOD FOR MICROBIOLOGICAL

INVESTIGATIONS BACKGROUND OF THE INVENTION Field of the Invention The present invention relates to a network-based method for the design and provision of test panels for a biological investigator.

Description of the Prior Art There are many microbiological studies that require an investigator to choose a panel of microorganisms to undergo some form of testing. Investigators must select such a panel in a variety of instances, as, for example, whenever a new anti-microbial agent is evaluated, whenever a "kill step" is designed for a food manufacturing process, whenever the prevalence of a newly-discovered gene is sought, or whenever a new diagnostic assay is developed.

In practice, several ad hoc procedures are available whereby organisms are selected by an investigator for a panel. The organisms may be selected randomly from a collection of a given species . Alternatively, the selection may be based upon whatever organisms are then available or upon prior experience. Care must also be exercised as to the number of organisms in the panel. If too small a panel is selected misleading conclusions may result because the chosen organisms may not be representative of a "real- world" population. On the other hand, a panel having too many organisms may also be disadvantageous since some of the organisms are likely to be genetically identical for practical purposes. This duplication can be inefficient and costly, especially when the experiment is inherently expensive.

U. S. Patent 5,660,981 (Grosz et al . ) discloses a method for the determination of diagnostic genetic markers for the identification of individual microorganisms. The method entails random amplification by polymorphic differences (RAPD) of the genomic DNA of a representative number of individuals from a genetically related population wherein that number will comprise a positive test panel. Similar amplification is performed on the genomic DNA of a significant number of individuals genetically unrelated to the positive test panel to comprise a negative test panel. No rationale is set forth explaining the selection of particular organisms. There is no teaching or suggestion as to a minimum adequate effective process useful to select the organisms for the categories of experiments conducted.

In view of the foregoing, it is believed advantageous to provide a method for the design of a panel using selection criteria that are rationally based on the immediate needs of the investigator so as to produce a panel that includes the minimum adequate number of organisms necessary to serve as an effective datum. It is also believed advantageous to provide a network-based service whereby the organisms selected for inclusion in a rationally designed panel may be provided to the investigator in a cost-effective manner .

SUMMARY OF THE INVENTION In a first aspect the invention is directed to a computer-implemented method for selecting candidate organisms for a biological screening panel. In response to an input identifying: i) the desired properties to be exhibited by the organisms; ii) the desired number (N) of organisms to be included in the panel; and iii) a choice of either a Common, Diverse or Narrow panel design strategy, and, if the panel design strategy is Narrow, the identification of a reference organism; a population is established by querying a database containing data records for a plurality of organisms to identify at least some organisms exhibiting the desired properties. The data record for all of the organisms in the population includes genetic typing information. The organisms in the population are clustered into Q number of groups based upon the genetic typing information of the organisms. N number of organisms are chosen from the Q groups in accordance with the selected panel design strategy, such that: i) where the panel design strategy is Common, the N organisms from the Q groups are chosen such that the largest of the Q groups is disproportionately over-represented in the N organisms; ii) where the panel design strategy is Narrow, each of the N organisms is chosen such that its genetic typing information is the same as or is a nearest neighbor to the reference organism; and iii) where the panel design strategy is Diverse, the N organisms are chosen so that each of the Q groups is represented in an approximately equal proportion among the N organisms . The choice of design strategy will be determined by the intended use of the panel.

The method may further include the step of modifying the composition of the panel by replacing one or more organisms originally included in the panel with a respective substitute organism, such that the relative proportion of the organisms in the panel having the desired properties is the substantially the same as the relative proportion of the organisms in the population having those desired properties.

In another aspect, the invention is directed to a network-based method for selecting and acquiring candidate organisms for a biological screening panel. After the customer-generated input is received over a computer network, the selection method as outlined above is implemented to produce a panel of organisms . The panel selections are presented over the network to the customer. In response to a customer authorization at least one organism is procured by locating a source of that organism in a database and sending a purchase requisition to that source.

The method is implemented in the context of a computer network, such as the Internet, interconnecting the customer and a server computer having available to it a computing resource for practicing the selection method.

BRIEF DESCRIPTION OF THE DRAWINGS The invention will be more fully understood from the following detailed description, taken in connection with the accompanying drawings, which form a part of this application and in which:

Figure 1 is a block diagram of a network-based system for selecting and providing organisms for a biological screening panel to a customer in accordance with one aspect of the present invention;

Figure 2 is a high level block diagram of a computer-implemented method for selecting organisms for a biological screening panel in accordance with another aspect of the present invention; Figure 3 is a histogram showing the relative frequency of genetic types for an example set of organisms; and

Figures 4-11 are successive screens as seen by a customer in an Internet-based implementation of the method of the present invention.

DETAILED DESCRIPTION OF THE INVENTION Throughout the following detailed description similar reference numerals refer to similar elements in all Figures of the drawings. Figure 1 is a block diagram of a system, generally indicated by the reference numeral 10, for implementing a method for selecting and acquiring candidate organisms for use in a biological screening panel in accordance with a first aspect of the present invention. The system 10 includes a user computer 12 and a server computer 1 . A customer at the user computer 12 may be any investigator desirous of obtaining a selection of candidate organisms (e.g., bacteria, fungi, algae, parasites) for a biological screening panel.

The user computer 12 and the server computer 14 are interconnected in an interactive manner by a computer network 16, such as the Internet. It should therefore be understood that, although not expressly illustrated, the user computer 12 has the requisite browser software and any suitable arrangement for effecting a connection to the network 16 (e.g., a telephone or cable modem connection to an internet service provider, an Ethernet connection, an area network connection to a server, or a wireless connection) . Similarly, although also not expressly illustrated, the server computer 14 also has access to the network 16 and has appropriate software for interacting with the user computer 12 in the manner to be described fully herein.

The server computer 14 has a computing resource 18 available to it for executing a program for implementing the rational panel design ("RPD") method for selecting organisms for a biological screening panel in accordance with another aspect of the present invention. The computing resource 18 may be located at any convenient location with respect to the server computer 16 and is connected thereto through an appropriate connection 20.

The system 10 also includes at least one, but more preferably, a plurality, of informational database (s) available to the server computer 14 and the computing resource 18. An available database, e.g. the database 22, may be viewed as "local" to the server computer 14 in the sense that it is accessible by the server computer 14 through a connection 24 and not through the network 16. Additionally or alternatively, the system 10 may include one or more information database (s) 26 that are available to the server computer 14 and the computing resource 18 through the network 16. The database (s) 22, 26 serve as repositories for data records for various strains of microbial organisms. The data records within the database (s) 22, 26 include extensive "biographical information" regarding properties exhibited by each strain of organism.

For example, within the database (s) 22, 26 an organism may be broadly classified in accordance the field of activity in which it is naturally encountered. Representative fields of activity include food, chemical and environmental. The database (s) 22, 26 may contain information as to the source from which the organism is isolated, such as whether the organism is a clinical isolate, a food isolate, an environmental isolate, or a soil isolate. Another property of interest for clinical isolates is the organ system affected by the organism (infection types), i.e., whether a given organism affects the lower respiratory system, the upper respiratory system, wounds, the urinary tract, or the gastrointestinal tract The source of infection, that is, whether the organism may be acquired nosocomially or through community acquisition, is also a useful property. Other interesting properties include the geographical location from which the isolate is derived, and, in the case of clinical isolates, the age and gender of the patient. The resistance of the organism to certain drugs, such as various antibiotics, is also an important property.

As will be developed, these and other properties exhibited by each organism serve as criteria whereby a given organism may be differentially selected over another organism.

Most significantly, the data record for the organisms in the database (s) 22, 26 includes genetic typing information. "Genetic typing information" is that information capable of dividing microbial organisms within a species into several groups based on their genetic similarity. All of the database (s) 22, 26 must employ equivalent genetic typing methodology so that meaningful comparisons may be made. Genetic typing information is discussed in Bergey's Manual of Systemic Bacteriology, Volume 1, 1984, ISBN 0-683- 04108-8, Noel R. Kreig (Editor), John G. Holt (editor- in-chief) , Williams & Wilkins, Baltimore, Maryland.

Genetic typing information may be obtained from typing methodologies such as the ribotype data gathered within and by DuPont Qualicon, Inc., Wilmington, Delaware, in association with the RiboPrinter^® microbial characterization system. Other sources of genetic typing information include 16S sequencing, Pulsed-Field Gel Electrophoresis (PFGE) , Multi-Locus Sequence Typing (MLST) , and suitably prepared DNA microarrays.

The server computer 14 is also interconnected through the network 16 to one or more culture collections 28 which serve a commercial source of organisms. An example of a culture collection is the American Type Culture Collection ("ATCC"), Rockville, Maryland. A general overview of the operation of the system 10 when implementing a network-based method for selecting candidate organisms for a biological screening panel in accordance with one aspect of the present invention is understandable from Figure 1. The information flow in the system 10 originates with a customer at the user computer 12. As represented by the reference character arrow 34 the initiating action is the transmission over the network 16 of a customer-generated request from the user computer 12 to the server computer 14. The request contains information concerning: i) the desired properties to be exhibited by the organisms; ii) the desired number (N) of organisms to be included in the panel; and iii) a choice of a panel design strategy (and a reference organism, if needed) . Upon receipt by the server 14, and in response to the customer-generated request, the resource 18 executes the rational panel design ("RPD") method for selecting organisms for a biological screening panel in accordance with another aspect of the present invention. The RPD method is explained in full detail in connection with Figures 2 and 3: As a result of the RPD method a number N of organisms exhibiting the desired properties is selected in accordance with the selected panel design strategy, based upon their genetic typing information.

In implementing the RPD method the resource 18 utilizes the database (s) 22, 26, as appropriate. After a panel is designed the results are presented over the network 16 to the customer at the user computer 12, as represented by the reference arrow 38.

Either with the original request or upon presentation of the results the customer may authorize procurement of the organisms selected. The authorization is transmitted over the network 16, as represented by the reference arrow 40. In response to a customer authorization a purchase requisition for one or more organisms is transmitted over the network 16 to at least one source 28. This action is represented by the reference arrow 42. The selection of a source for a needed strain of organism is based upon factors such as proximity and cost to the customer. The fulfillment of the order is sent from the source 28 to the customer, as indicated by the arrow 44.

The customer is billed for the strains of organisms purchased in addition to the service in providing a rational panel design for selecting organisms for the panel. Figure 2 is a high level block diagram of the steps of the computer-implemented rational panel design

("RPD") method implemented by the computing resource 18

(Figure 1) to select candidate organisms in accordance with the present invention.

As indicated by block 48 information regarding the organisms to be included in the panel is received as an input by the computing resource 18. It will be recalled that the customer-generated request 34 (Figure 1) specifies the desired number (N) of organisms to be included in the panel and the desired properties to be exhibited by the included organisms. Any of the above- discussed properties exhibited by an organism may be specified by the customer. In addition, the customer selects either a Common, Diverse or Narrow panel design strategy. As a proviso, if the customer selects a Narrow panel design strategy, a reference organism is also identified. In some instances the customer- generated request may also include the specification of a number R, for a purpose to be explained.

As will be more fully developed use of a Common panel design strategy results in a panel containing a set of organisms that is representative of the most common genotypes having the desired properties . With a Narrow panel design strategy the panel contains a distribution around the identified reference organism. Under a Diverse panel design strategy a panel is created that contains as many different genetic types as possible. In response to the customer-generated request, as indicated at block 50, one or more of the database (s) 22, 26 containing data records for a plurality of strains of organisms is (are) queried to identify at least some organisms exhibiting the desired properties . The identified organisms establish a population on which further selection actions are taken. Organisms not exhibiting the desired properties are excluded from further analysis. As noted, the data record in the databases for each of the organisms in the population includes genetic typing information regarding the organism.

Next, the organisms in the population are clustered into Q number of groups. This step is indicated at block 52. The clustering is based upon the genetic typing information of the organisms. Any suitable clustering algorithm may be utilized, such as the algorithms implemented in the program sold by Applied Maths BVBA under the trademark BioNumerics^®. Examples of clustering algorithms include the Leader Algorithm, the K-means Algorithm, the Single-linkage Algorithm, and the Two-way Joining Algorithm, as described in J. A. Hartigan, Clustering Algorithms, John Wiley & Sons, Inc., New York (1975).

For purposes of illustration and further discussion of the method of the present invention a typical histogram resulting from the application of the clustering step is indicated in Figure 3. The histogram is based upon a data depicted in a poster

"Geographical Distribution and Antibiotic Resistance of Clinically-Significant Staphylococcus aureus Ribotypes as Determined by an Electronically-Linked Network of Automated Ribotyping Systems, 1999, M.A. Pfaller, et al . , Abstract, 99th General Meeting, Am. Soc .

Microbiol . Abstr. L/U-2, pg. 435, American Society for Microbiol., Washington, D.C.

The population is clustered into eleven groups (i.e., Q is eleven) identified by letters A through K along the horizontal axis on the basis of a genetic typing method known as ribotyping, and the genetic types are therefore' designated as "ribotypes". The count along the vertical axis of the histogram shows the number of ribotypes in each group, while the histogram itself reflects the relative frequency of each of the ribotype in the population. The count of each ribotype, and the strain identifier for each group (keyed to the Pfaller poster) are set forth in Table 1 as follows :

TABLE 1

Group Count Identifier

A 118 105-694-S-3

B 97 105-672-S-4

C 65 130-189-S-4

D 35 105-684-S-3

E 28 105-672-S-5

F 19 105-671-S-4

G 15 105-685-S-5

H 12 105-708-S-8

I 7 105-686-S-5

J 3 130-191-S-5

K 1 Any of several

The total number of organisms in the population is four hundred (400) . In the example that will be developed, it is assumed that the desired number (N) of organisms to be included in the panel is seventy-five (75) organisms.

As indicated at decision block 54 the selection of the number N of organisms from the Q groups that will comprise the panel is effected in accordance with the selected panel design strategy.

Common Panel Design Strategy Where the panel design strategy is Common (block 56) the N organisms from the Q groups are chosen so that the larger groups are more heavily represented in the panel.

For example, in one implementation of a Common panel design strategy, if N is much smaller than the total number of organisms in the population, the N organisms may be chosen so that the relative proportion of the N organisms having membership in a given group (i.e., having a given set of genetic typing information) is the same as the relative proportion of organisms having membership in that group in the population. Since a panel cannot contain a fraction of a strain, only representatives of the most commonly- occurring genetic types will be included in the panel. The Common design strategy may be used to obtain typical representatives of a certain microbial class for further investigation. For example, if one wished to obtain a set .of antibiotic-resistant bacterial strains for gene sequencing in order to understand the general resistance mechanisms which applied to that species, one would use a Common panel design strategy.

In the context of the example depicted by the histogram of Figure 3 the relative percentage of the each group within the population and the number of organisms selected from that group under a Common panel design strategy is set forth in the following Table 2.

TABLE 2

Relative

Number

Group ount Percentage Selected

A 118 29.5 22 B 97 24.3 18 C 65 16.3 12 D 35 8.8 6 E 28 7.0 5 F 19 4.8 4 G 15 3.8 3 H 12 3.0 2 I 7 1.8 1 J 3 0.8 1 K 1 0.3 1

An alternative implementation for selecting the N organisms in accordance with a Common panel design strategy the N organisms would be chosen at random from among only the first R groups, where R < Q. As noted the number R is derived from the customer-generated request. (The numbers N and R may be equal to each other.) For example, one could choose the N organisms from only the two or three largest groups .

Narrow Panel Design Strategy Where the panel design is Narrow (block 58), each of the N organisms in the panel is chosen so that its genetic typing information is the substantially the same as or is a nearest neighbor to the reference organism. The Narrow design strategy would be used to obtain microorganisms similar to a designated strain. Using a Narrow panel an experimenter can get the closest genetic relatives of a given strain. For example, if one had discovered a strain which produced a useful biological product and one wished to screen other strains to see if they would produce the same product in higher yield, one would use a Narrow panel design strategy.

The determination of the various degrees of genetic proximity to the reference organism may be accomplished using any nearest genetic neighbor identification algorithm. The above-identified BioNumerics^® program sold by Applied Maths BVBA under the trademark BioNumerics^® includes such an algorithm. The software provided with the RiboPrinter^® Microbial Characterization System (DuPont Qualicon, Wilmington, DE USA) is also capable of sorting a list of organisms in order according to their genetic similarity to a given organism, as determined by ribotyping. Thus, the N nearest neighbors will be the first N strains in the sorted list.

Diverse Panel .Design Strategy Where the panel design is Diverse (block 60) , the N organisms are chosen such that all Q groups are represented in the panel in approximately equal numbers . The Diverse design strategy is used when it is necessary to ensure that all genetic varieties within a given population are covered. For example, if one wished to evaluate a diagnostic test which claimed to detect a given species, one would perform the evaluation using a panel created with the Diverse panel design strategy. The' Diverse panel design strategy would also be used to validate that a process step was capable of killing all bacteria of a certain kind.

In one instance the selection of a diverse panel can be accomplished by cycling through the Q groups, starting with the most populous and moving sequentially toward the least populous group, and selecting an organism from each group. Selection continues until N organisms are selected. Selection occurs with the proviso that once any group is depleted it is dropped from the cycle. Groups with the same number of organisms are ordered arbitrarily into the sequence. A Diverse panel may thus be implemented in the context of the example of Figure 3 as shown in Table 3 (generally similar to Table 2 save for the entries in the "Number Selected" column) .

TABLE 3

Relative Number . Group Count Percentage Selected

A 118 29.5 8

B 97 24.3 8

C 65 16.3 8

D 35 8.8 8 E 28 7.0 8

F 19 4.8 8

G 15 3.8 8

H 12 3.0 8

I 7 1.8 7 J 3 0.8 3

K 1 0.3 1

An alternative implementation for selecting a diverse panel would be to select one of the Q groups at random, and then select an organism from that group, and repeat this process until N organisms had been selected. A group is skipped if all its members are already chosen for a panel. As a further optional refinement, by whatever strategy the panel is selected, the composition of the panel may be modified (Block 62) by replacing one or more organisms originally included in the panel with a respective substitute organism, such that the relative proportion of the organisms in the finally constituted panel that have the desired properties is substantially the same as the relative proportion of the organisms in the population having those desired properties. This may be done by selecting that set of organisms having the desired properties for inclusion in the panel (as identified in the customer-generated request) which minimizes a chi-squared statistic comparing the panel to the population with respect to the user-selected variables. Information also may be presented about the relative frequency of each genotype in the population.

In the context of the example under discussion, suppose it is desired for purposes of the experiment to match the panel (by whatever panel design strategy chosen) to the population on the basis of patient gender. Suppose that in the entire population of four hundred (400) isolates there were one hundred thirty- three (133) from men and two hundred sixty-seven (267) from women. When organisms are chosen for the panel, the organisms would be selected, assuming sufficient organisms existed in each group, so as to give the same proportion. In other words, twenty-five (25) organisms from men and fifty (50) from women.

As discussed in connection with Figure 1 the N organisms finally selected by the rational panel design method to constitute the panel are presented over the network 16 to the customer.

Figures 4 through 11 show successive screen displays presented to a customer to impart a sense of the "look and feel" of the implementation of the present invention in the environment of a computer network, such as the Internet. It should be understood that the screens presented are preferred but may be readily varied by one of skill in the art without departing from the scope of the invention. The screen displays are programmed in Hypertext Markup Language (HTML) . Figure 4 is the introductory log-in screen 100 that a customer sees upon log-in to the Internet web site hosted by the server computer 14. Screen 100 provides a common "username/password" log-in procedure for a prior customer. If the customer is a first-time user additional needed information (such as name, address, billing information) is sought in a registration screen, not shown.

Figure 5 shows the next screen 102. This screen 102 prompts the customer to supply pertinent information regarding the desired properties to be exhibited by the organisms in the panel.

The information is solicited in eleven fields . In field 104 the customer indicates isolate choice using a pull-down menu. The menu choices include Clinical Isolates, Food Isolates, Food-plant Isolates, Environmental Isolates, and Soil Isolates. In the example of Figure 5 "Clinical Isolates" has been chosen .

The organ system affected by the organism may be indicated by pull-down menu in field 106, indicated as "Infection Type" . The menu 106 includes Lower Respiratory, Upper Respiratory, Wound, Urinary Tract, Gastro-intestinal, Diarrheal, and "All Infection types" (shown as chosen) . Fields 108, 110 permit indication of the sources of infection as either or both

"Nosocomial" (useful in hospital epidemiology and the like) and/or "Community Acquired". Key entry fields 112, 114 are for the age range of the individual affected by the organism (with "20" to "30" being respectively entered) . Field 116 is a pull-down menu that permits selection of the gender affected, either "Male", as shown, or "Female". Field 118 is a pulldown menu used to indicate geographic constraints on the organism from a menu that includes "North America", "South America", "Europe", "Mid & Near East", "Asia Pacific", or "Not Specified". Field 120 permits a country designation to be entered by way of a pull-down menu that includes the roster of the United Nations and "Not Specified". The "United States" is highlighted. If desired, the "Genus" and "Species" identities of an organism may be input through fields 122, 124 respectively. Pull-down menus list all of the genus/species available in the database (s) . In this example the further narrowing of the "Species" to "aureus" is consistent with the genus "Staphylococcus" selected in the field 124.

If drug susceptibility is also a desired property of the organism, information regarding resistance to various drugs will be permitted to be entered in pulldown menu fields 125A through 125E in Figure 7. Each of these fields 125A through 125E offer choices of "susceptible", "intermediate", "resistant" and "not specified" . The specific drugs displayed for selection in the fields 125A through 125E depend upon the genus and species selected in fields 122, 124. Thus, fields 125A through 125E will specify drugs Oxacillin, Erythromycin, Tetracycline, Clindamycin and Vancomycin which affect "Staphylococcus" (field 124) and "aureus" (field 124) . A different array of drugs would be presented for a different genus and species.

As shown in Figure 6, in screen 126, the desired number (N) of organisms to be included in the panel is input by the customer in field 128. The choice of a "Common", "Diverse" or "Narrow" panel design strategy is effected by pull-down menu in field 130. A brief explanation of each strategy is also given in this screen 126. If the "Narrow" panel design strategy is selected, the identity of a selected reference organism may be keyed into a field 131.

Screen 132 in Figure 7 shows the results of an initial querying of the database (s) 22, 24. Field 134 lists the number of records (i. e., the population) that matches specified criteria (other than antibiotic resistance) . The illustration shows a population of "5,532" in that field 134. If the population is too large, the genus and species selection may be narrowed using a scroll-down menu 136. The drug resistance properties are entered at fields 125A through 125E ("not specified" in the example illustrated) at this point . Bar graph in the field 138 and the table the field 140 respectively graphically and tabularly illustrate the selected drug resistance for the highlighted genus and species .

Turning now to Figure 8, the screen 142 shows in field 144 the population (776 strains in this example) which meets the additional user-specified criteria of antibiotic resistance as specified in the fields 125A through 125E (Figure 7) . Also shown in field 146 is a histogram (similar to that shown and discussed in connection with Figure 3) that show clusters of organisms based on their genetic typing information.

In the next screen 148 on Figure 9, the result of rational panel design program for the chosen panel design strategy is shown. In this example the result for a "Common Panel" is presented to the customer, both as spotlighted by "X's" on the graph in the field 150 and as a list of specific isolates in columnar form in the field 152. The column 152 shows all strains chosen by the RPD program. Specific organisms may be selected using check boxes 154. The purchase price for each organism is set forth in the event the customer desires to purchase any of the listed organisms. Various activities ("Remove", "Keep" "View Details" and "Purchase") may be selected using buttons 156. In Figure 10 the screen 158 provides a field 160 in which the customer may apply an identifier to the selected panel. Also, the customer may use the newly- selected panel to update a previously saved panel. The previously saved panel is selected using a scroll down menu 162. The resulting selection is saved using a button 164.

The screen 166, shown in Figure 11, summarizes the desired transaction. On this screen the isolates selected for purchase are listed in columnar form with purchasing information (such as isolate number, label, source, species name and price) set forth. Check-out button 168 completes the transaction and executes the purchase order.

It will be seen that a person of skill in the biological sciences who requires a test panel against which research or validation data can be compared can, by using the method of the invention, have access to an interactive, rational, panel-design system over a computer network without any human intervention other than by the user. Moreover the database, itself networked, is larger and more comprehensive than most workers in the field have at hand or can reasonably assemble without extensive research. In addition, this database is under constant update. The customer, working at an interface remote from the data processing system, is connected through a communications network to a specialized data processing system that will design a panel tailored to the asserted needs of the user. Further, any one or all of the components needed to assemble the designed panel can be ordered on-line from one source despite the fact that those components may be available from many diverse locations worldwide. The economic advantages of the method are considerable compared to random, intuition-based, panel design without genetic-typing information usually with attempts to ensure adequacy by sheer numbers . Eliminating multiple occurrences of nearly-clonal isolates can save perhaps eighty percent (80%) on screening costs. Those skilled in the art, having the teachings of the present invention as hereinbefore set forth may effect numerous modifications thereto. It should be appreciated that such modifications are to be construed within the contemplation of the present invention, as defined by the appended claims.

Claims

What is claimed is :

1. A network-based method for selecting candidate organisms for a biological screening panel, the method comprising the steps of: a) receiving over the network a customer-generated request containing information concerning: i) desired properties to be exhibited by the organisms, . ii) the desired number (N) of organisms to be included in the panel, and iii) a choice of a panel design strategy; b) in response to the request establishing a population of organisms exhibiting the desired properties, the data record for all of the organisms in the population including genetic typing information; c) based upon the genetic typing information of the organisms, selecting the number N of organisms in accordance with the selected panel design strategy; and d) presenting over the network the panel selections to the customer.

2. The method of claim 1 further comprising the step: e) in response to a customer authorization, sending a purchase requisition to a source for at least one of the selected organisms.

3. A network-based method for selecting candidate organisms for a biological screening panel, the method comprising the steps of: a) receiving over the network a customer- generated request containing information concerning: i) desired properties to be exhibited by the organisms, ii) the desired number (N) of organisms to be included in the panel, and iii) a choice of either a Common, Narrow or Diverse panel design strategy, provided that, if the panel design strategy is Narrow, a reference organism is also identified; b) in response to the request, querying a database containing data records for a plurality of organisms to identify at least some organisms exhibiting the desired properties, thereby to establish a population, the data record for all of the organisms in the population including genetic typing information; c) clustering the organisms in the population into Q number of groups based upon the genetic typing information of the organisms; d) selecting, in accordance with the selected panel design strategy, a panel containing the number N organisms, such that: i) where the panel design strategy is Common, the N organisms from the Q groups are chosen such that the largest of the Q groups are disproportionately over-represented in the N organisms; ii) where the panel design strategy is Narrow, each of the N organisms is chosen so that its genetic typing information is the substantially the same as or is a nearest neighbor to the reference organism; and iii) where the panel design strategy is Diverse, the N organisms are chosen so that each of the Q groups is represented in an approximately equal proportion among the N organisms; and e) presenting over the network the panel selections to the customer.

4. The method of claim 3 wherein the database is connected to the network.

5. The method of claim 3 further comprising the step of: f) in response to a customer authorization, procuring at least one organism by locating a source of that organism in a database and sending a purchase requisition to that source.

6. The method of claim 3, wherein after selection of the panel in step d) , the method further comprises the step of: modifying the composition of the panel by replacing one or more organisms originally included in the panel with a respective substitute organism, such that the relative proportion of the organisms in the panel having the desired properties is the substantially the same as the relative proportion of the organisms in the population having those desired properties .

7. The method of claim 3, wherein, when the panel design strategy is Common, the N organisms from the Q groups are chosen so that the relative proportion of the N organisms having membership in a given group is the same as the relative proportion of organisms having membership in that group in the population.

8. The method of claim 3, wherein, when the panel design strategy is Common, the N organisms from the Q groups are chosen at random from among only the first R groups, where R < Q.

9. The method of claim 8, wherein the customer- generated request contains information concerning the number R .

10. The method of claim 3, when the panel design strategy is Diverse, the N organisms are chosen in step d) by dl) cycling through the Q groups, moving sequentially from the most populous toward the least populous group, and d2) selecting an organism from each group until N organisms are selected, with the proviso that any depleted group is dropped from the cycle.

11. The method of claim 3, when the panel design strategy is Diverse and the smallest of the Q group contains at least N/Q organisms, the N organisms are chosen in step d) by dl) selecting one of the Q groups at random and selecting an organism from that group, and d2 ) repeating step dl) until N strains had been selected.

12. A network-based method for selecting candidate organisms for a biological screening panel, the method comprising the steps of: a) transmitting over the network a customer- generated request containing information concerning: i) desired properties to be exhibited by the organisms, ii) the desired number (N) of organisms to be included in the panel, and iii) a choice of either a Common, Narrow or Diverse panel design strategy, provided tha t, if the panel design strategy is Narrow, a reference organism is also identified; b) receiving over the network a panel that is selected by: i) in response to the customer- generated request, querying a database containing data records for a plurality of organisms to identify at least some organisms exhibiting the desired properties, thereby to establish a population, the data record for all of the organisms in the population including genetic typing information; ii) clustering the organisms in the population into Q number of groups based upon the genetic typing information of the organisms; iii) selecting, in accordance with the selected panel design strategy, a panel containing the number N organisms, such that:

A) where the panel design strategy is Common, the N organisms from the Q groups are chosen such that the largest of the Q groups are disproportionately over- represented in the N organisms;

B) where the panel design strategy is Narrow, each of the N organisms is chosen so that its genetic typing information is the substantially the same as or is a nearest neighbor to the reference organism; and

C) where the panel design strategy is Diverse, the N organisms are chosen so that each of the Q groups is represented in an approximately equal proportion among the N organisms .

13. The method of claim 12 further comprising the step of: c) transmitting over the network an authorization to procure at least one organism.

14. A method for selecting candidate organisms for a biological screening panel comprising the steps of: a) in response to an input identifying i) desired properties to be exhibited by the organisms, ii) the desired number (N) of organisms to be included in the panel, and iii) a choice of either a Common, Diverse or Narrow panel design strategy, provided that, if the panel design strategy is Narrow a reference organism is also identified, querying a database containing data records for a plurality of organisms to identify at least some organisms exhibiting the desired properties, thereby to establish a population, the data record for all of the organisms in the population including genetic typing information; b) clustering the organisms in the population into Q number of groups based upon the genetic typing information of the organisms; c) selecting from the Q groups the number N of organisms in accordance with the selected panel design strategy, such that: i) where the panel design strategy is Common, the N organisms from the Q groups are chosen such that the largest of the Q groups are disproportionately over-represented in the N organisms; ii) where the panel design strategy is Narrow, each of the N organisms is chosen such that its genetic typing information is the same as or is a nearest neighbor to the reference organism; and iii) where the panel design strategy is Diverse, the N organisms are chosen so that each of the Q groups is represented in an approximately equal proportion among the N organisms.

15. The method of claim 14, wherein after selection of the panel in step c) , the method further comprises the step of: d) modifying the composition of the panel by replacing one or more organisms originally included in the panel with a respective substitute organism, such that the relative proportion of the organisms in the panel having the desired properties is the substantially the same as the relative proportion of the organisms in the population having those desired properties .

16. The method of claim 14, when the panel design strategy is Common, the N organisms from the Q groups are chosen so that the relative proportion of the N organisms having membership in a given group is the same as the relative proportion of organisms having membership in that group in the population.

17. The method of claim 14, wherein, when the panel design strategy is Common, the N organisms from the Q groups are chosen at random from among only the first R groups, where R < Q.

18. The method of claim 17, wherein the customer- generated request contains information concerning the number R.

19. The method of claim 14, when the panel design strategy is Diverse, the N organisms are chosen in step d) by dl) cycling through the Q groups, moving sequentially from the most populous toward the least populous group, and d2 ) selecting an organism from each group until N organisms are selected, with the proviso that any depleted group is dropped from the cycle.

20. The method of claim 14, when the panel design strategy is Diverse and the smallest of the Q group contains at least N/Q organisms, the N organisms are chosen in step d) by dl) selecting one of the Q groups at random and selecting an organism from that group, and d2 ) repeating step dl) until N strains had been selected.