WHY WE CLASSIFY

We put labels on groups of similar concepts for convenience of thinking about them. We labels words as nouns, verbs, and so forth. We belong to clans, tribes, religions, races, countries, and regions to which we give names. Classes of life forms are classified in two main frameworks. First, we classify them based on their history (phylogenetic groups) along with what they are capable of doing (phenetic groups). Second, we classify them with respect to their relationships to their environment and other life forms within that environment (ecologic groups).

CLASSIFYING LIVING ORGANISMS

The first framework is established without regard to habitat. That is, a classification based solely on genetic or phenetic characterization is possible on organisms without regard to their source. Such classification can be accomplished solely with organisms from culture collections. The classification requires no knowledge of source or f source or habitat.

To classify organisms in terms of ecology depends absolutely on knowledge of where the organisms normally live. Ecologic studies can be performed on organisms from culture collections, however the information on the source of the organisms must be available. The more detailed the source information given, the more precise the study. The source information will require interdisciplinary communication. For example, many modern ecologic studies use advanced geography methods (Geographical Information Systems or GIS) to describe the place of sampling.

As detailed later, the classes in either framework, taken together, describe aspects of diversity. Clearly, culture collections are repositories for diversity information as well as the organisms themselves. Further, culture collections are repositories containing much of the working material for biotechnology.

INTERDISCIPLINARY COMMUNICATION

Consider the persons and disciplines involved in devellved in developing such processes: the molecular biologist models the insertion of the new gene, the biologist finds an organism containing the gene, the chemist predicts the structure and function of the product, the immunologist finds epitopes for antibody adsorption to purify the product, the microbiologist projects the new growth rate, morphology, and biochemistry of the host strain, the regulator predicts hazard, and the ecologist predicts long term interactions. Each of these disciplines have their own vocabulary and conventions of communication. Modern biotechnology clearly requires interdisciplinary communication.

Even within a discipline such as microbiology, communication problems exist. The most common way of describing an organism is through the species name. Thus, we "identify" an organism as belonging to a particular species. However, different methods such as pure culture phenotypic assessment methods can give very different identifications from immunologic or genetic probes. This problThis problem becomes critical if we wish to follow changes in the microbial populations of two sites by different methods.

Consider the following example where the same samples from human periodontal pockets were sampled. The first column of the table presents the percentage of samples wherein each organism was identified and scored as present by both the immunologic probes and by traditional pure culture techniques. The next two columns give the percentage of samples wherein the methods disagreed. The last column gives the percentage of times the two methods detected the organism divided by the positive agreements plus the disagreements (col. 1/col. 1 + col. 2 + col. 3).

COMPARISON BETWEEN A "PROBE" METHOD AND CULTURE

  The agreement ranges from none to a high of 81%. Conclusions based on incidence of these organisms in the sampled sites will vary depending on the method of detection used. This is a clear example of problems of communication that exist even in closely related branches of microbiology, immunologic assay and pure culture microbiology.

 The limiting factor in interdisciplinary communication is the ability of the persons involved to understand the vocabulary and nomenclature of all the disciplines included to the level needed to reason with the information. The user of information must receive input from disparate disciplines each of which have their own jargon. The practitioners of the various disciplines invent new words and give new meaning to oldeaning to old words as their disciplines advance. This inevitable practice facilitates communication locally but creates noise globally.

BUILDING A DICTIONARY FOR MICROBIOLOGY

We would like to build a dictionary for each discipline, i.e., a standard nomenclature and vocabulary. Our immediate field is straightforward. We are comfortable with the words we use. However, we have problems using the standards of other disciplines.

 Illustrative of these points is the task of building a standard nomenclature and a vocabulary to describe species of microorganisms. The building of such a dictionary has two distinct aspects: the names of organisms and the attributes of organisms. They must follow different sets of rules. We shall consider the "dictionary" for attributes of organisms in a separate discussion.

MICROBIAL NOMENCLATURE

Perhaps the most pervasive element in biological databases is, and will continue to be, the "name" given to the" given to the taxonomic place of the organism. Usually this will revolve around the species name. Within each of the major groups of biological entities good standardization of rules of nomenclature exits. The rules for each group are maintained by the several nomenclatural bodies set up within two of the Unions adhering to International Council of Scientific Unions: the International Union of Biological Societies and the International Union of Microbiological Societies.

The rules for naming the various major groups of microorganisms fall under different Codes of Nomenclature. The code for viruses is managed by the International Committee for the Taxonomy of Viruses. The fungi and algae are covered by the Botanical Code. With some redundancy, the Zoological Code covers algae and protozoa. Such organisms as the cellular slime molds, which are amoebae that differentiate into fungal appearing forms also are covered in both codes.

The bacteria have their own International Code of Bacterial Nomenerial Nomenclature. The rules of this code are promulgated by the international Committee on Systematic Bacteriology. A major distinction between the Bacteriological Code and the others is that all names of bacteria were reset by the publication of the Approved List of Valid Names in 1980. Additions and emendations must be published in the International Journal Of Systematic Bacteriology (IJSB). All new taxa must be listed in the IJSB for validity, regardless of where the original description appeared. Thus, only one journal need be consulted for valid names of bacteria and only from 1980 onward.

The following outlines the existing codes accepted in microbiology:

=> CODES OF NOMENCLATURE

VIRUSES => ICTV (IUMS)

FUNGI, ALGAE => BOTANICAL CODE (IUBS)

PROTOZOA, ALGAE => ZOOLOGICAL CODE (IUBS)

BACTERIA => BACTERIOLOGICAL CODE (IUMS)

The rules allow scientific s allow scientific conflict. Unfortunately for the database producer, a clear set of authoritative names is not

available from the nomenclaturists in most cases. With the exception of the International Committee on Taxonomy of Viruses, the committees responsible for establishing the naming rules do not judge scientific validity of the names, only the correct etymology in some cases.

A continuing problem faced by database and culture collection catalogue producers is the lack of availability of such "official" lists in publicly-accessible, machine readable form. The publishers of these lists (and many other lists of standards) often maintain these lists in computer files. However, the prevailing view of these publishers is that they wish to sell hard copy (books or journals) rather than have people make and use electronic copies. In some cases, these lists have been retyped into databases and made available to the public as value-added databases.

ECOLOGICAL CULTURE COLLULTURE COLLECTING

There are many reasons for establishing a culture collection. Most have an ecological motivation of some sort. The major exception is simply obtaining cultures from another collection. In this context, we can consider medical collections as ecologically based since the ecology involved is that of the host organism. To start a biotechnology collection the econiche where the organism with the desired properties is sampled and candidate organisms accessioned into the collection. The following scheme illustrates the iterative nature of establishing a culture collection from nature and classifying the organisms thus obtained.

ESTABLISHING STRAIN POPULATION FOR ACCESSION AND STUDY

• sample econiche to classify predominant biotypes by morphology, Gram stain, tolerance and use of oxygen
• isolate and accumulate representatives of each major biotype
• assign biotypes to appropriate taxa
• establish minimum blish minimum phenetic span for inclusion into each biotype

DATA GATHERING ON ISOLATES IN EACH CATEGORY

• collect data on physiology, biochemistry, inhibitory tolerance, morphology, serology, DNA homology, etc. of isolates

DATA MANAGEMENT AND QUALITY CONTROL

• store and search data (simple bookkeeping)
• evaluate tests
• compare results of repeats as controls
• feed back results to laboratory
• generate reports

ANALYSIS

• compare by direct matching with authentic strains
• evaluate test for applicability and discriminatory power
• do cluster analysis (numerical taxonomy)
• determine feature frequencies
• choose taxonomy (by informational content for intended use)

CLASSIFY ISOLATES

• establishment of identification methods
• construct keys
• construct probabilistic matrix

As the collection grows, the above process is repeated and the taxonomy is updated. Why not use a "standard" taxonomy? As stated above, there is no "official" scientific definition of each taxon. Therefore, the individual microbiologist must use their best professional judgement to identify a new isolate. As we shall see, computers can help in this process.

The picture is murky when we consider the application of "standard test methods" to study of microbial populations and their member taxa. In almost all cases, the consideration of microbes in the environment falls victim to oversimplification. The foremost, but hardly only, oversimplification is to consider the microbes on the species level when it is a fundamental tenet of microbiology that phenotypic expressionotypic expression of a particular functional capacity is an isolate or strain-specific property. (While there is a conceptual difference between the "isolate" and "strain", discussed later, the distinction is not required for the immediate discussion.)

The practical importance of this over-generalization is that we extrapolate from one population to another, assuming the spectrum of organisms is applicable across these classes. Yet, microbial ecologists know very well that seemingly small changes in the harboring econiches yield organisms that are often different enough to be "unidentifiable" to species level. "Often" translates to 50-100% of the isolates.

The information required to study ecology in terms of both taxonomy and diversity may be outlined as follows. Each of the following categories of information become data elements in a database.

RECORDS REGARDING INDIVIDUAL ISOLATES

• Describe a placeescribe a place and time
• Sample the place
• Describe the organisms in the sample
• Classify the organisms in the sample
• Relate the classes to known organisms (identify)
• Name unknown organisms

The resulting data are entered into a computer. Many database management systems exist. Their internal logic of data storage varies considerably. However, the resulting database may be visualized as a simple two dimensional table no matter which system is used.

|Isolate|Sample|Taxon|History| Features|
| 1     |      |     |       | - - + + |
| 2     |      |     |       | + - + + |
| 3     |      |     |       | - - + - |
| .     |      |     |       | + - - - |
| N     |      |     |       | + - + - |

For ecologic purposes, we wish to ascertain properties of the sampleties of the sample population as a whole. The data to do so may fall into the following categories.

RECORDS REGARDING POPULATIONS SAMPLED

• Describe population distribution in sample
• Deduce capacity of population to effect change
• Relate organisms classes to dynamics of place

In these information gathering procedures, we should accept the reality of the following statements.

WE CANNOT:

• isolate & study phenotype of all organisms in a sample
• analyze all the properties of the organisms we can isolate
• relate macromolecular sequences in samples to phenotype of non-cultured organisms

WE CAN: classify & identify isolates