Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Chemical analysis
Reexamination Certificate
2001-05-09
2004-06-29
Hoff, Marc S. (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Chemical analysis
C435S287300
Reexamination Certificate
active
06757618
ABSTRACT:
TECHNICAL FIELD
The present invention relates to chemical structures and, more particularly, to a method and apparatus for assigning numeric or alpha-numeric identifiers to chemical structures and for identifying chemical equivalency.
BACKGROUND ART
In the field of chemistry there are numerous molecular properties that are of possible interest to chemists. Molecules may be compared for equivalency based on these properties. For example, a high level or general comparison of equivalency may compare two compounds or molecules with respect to the number of non-hydrogen atoms they contain or the number of bonds they have. Alternatively, a more detailed comparison of equivalency may compare chemical structures based not only on the number of bonds or atoms, but on specific atomic connectivity or spatial relationships of those atoms. For example, ring systems or cyclic systems of two molecules or compounds may be compared for equivalency.
A cyclic system is a chemical structure in which atoms are bonded together to form single or multiple rings. Cyclic system equivalence is of particular interest in the field of medicinal chemistry where the equivalency of cyclic systems may correspond to physiological or biological properties. However, a particular criterion of molecular equivalence only becomes operable, in a practical sense, when a chemist can identify each of the distinct classes of molecules that are equivalent with respect to the properties of interest.
Relative and absolute identification schemes are commonly used for identifying chemical structures that are molecularly equivalent. Relative identification schemes assign a unique identifier to each molecular structure encountered by the identification scheme. The assigned identifiers are not related to any particular information in a chemical structure. For example, a relative identification scheme may assign an identifier of one to a first chemical structure, an identifier of two to a second chemical structure and so on. A relative identification scheme, therefore, requires a memory that stores a list of identifiers that have been previously assigned to molecular structures.
An absolute identification scheme assigns an identifier to a molecular structure based solely on the information available in the molecular structure being identified. For example, an absolute identification scheme may assign a chemical structure having three atoms and two bonds, the identifier
32
, wherein the first digit represents the number of atoms in the structure and the second digit represents the number of bonds in the structure. Absolute identification schemes are beneficial in that the scheme need not check to see if an identifier is in use when assigning a new identifier. Additionally, through the use of an absolute identification system, two collections of compounds (e.g., molecular structures) may be directly compared with respect to the molecules they contain without coordinating their identifiers.
Criteria of equivalence are useful when a chemist is selecting compounds (e.g., collections or mixtures of molecules) for purchase. When selecting compounds for purchase, the chemist may first filter the list of compounds to screen out the compounds that are clearly of no interest. After screening out the uninteresting compounds, the chemist may visually inspect the remaining compounds. The chemist may sort the remaining compounds according to their cyclic system identifiers and, therefore, may include or exclude portions of compounds having common cyclic system identifiers. In selecting compounds for purchase, the use of cyclic system identification may save time and reduce error in selecting compounds for purchase. However, if a particular identification system erroneously assigns the same cyclic system identifier to compounds having different cyclic systems, the chemist loses faith in the fidelity of the identification system and the ability of the identification system to distinguish different chemical structures.
Criteria of equivalence are also useful in comparing two or more different collections of compounds. For example, if a chemist desires to know which compounds are similar with regard to particular properties among compound collections and which compounds differ with regard to particular properties among the compound collections, the chemist may use an identification system to name or identify each compound in the two compound collections. If the identification system the chemist uses is an absolute identification system, the chemist may simply compare the identifiers of the chemical structures of the compounds in the two compound collections. An identifier common to the compound collections indicates a common compound between the compound collections. A unique identifier in one of the collections indicates a compound found only in that collection.
In chemical and drug research, chemists often construct compound screening collections or libraries. Screening collections are used to scan a subset of a collection of compounds for a particular activity, rather than scanning the entire compound collection. The subset could be designed to emphasize particular types of compounds or could be designed to contain dissimilar compounds. If the cyclic systems of the compounds are used as a typing criterion for the collection, screening subsets are easily constructed. For example, after the chemist uses a filtering process to exclude compounds that the chemist does not wish to consider, the chemist may randomly order the cyclic-system identifiers and then select the number of compounds the chemist wishes from each successive cyclic system group until the chemist has a subset of the desired size.
If a screening operation has a large number of active compounds (compounds active in a biological test system of interest to a project team), the task of focusing on which of those compounds (called leads or hits) to pursue as useful starting points for lead optimization can be difficult. Numerous factors enter into the evaluation of a lead and, in many cases, close analogs of an active compound exist which differ at only one position by a small structural change from the active compound. A structure activity relationship (SAR) is sometimes said to exist if a chemist finds pairs of close analogs that differ significantly in their activity. Grouping compounds by cyclic system greatly accelerates and systematizes the process of finding such pairs. Finding such an SAR supports the choice of that cyclic system for one criterion to be used in finding leads.
In lead optimization efforts, large numbers of closely related compounds may be synthesized and tested. These efforts are guided by a growing understanding of the related SARs. SARs evolve out of numerous pairwise comparisons of closely related structures. If N compounds related to a lead exist there are N(N−1)/2, or roughly N
2
/2, possible pairwise comparisons that must be considered. If N is between 1,000 and 10,000, there may be between roughly 500,000 and 50,000,000 pairwise comparisons.
Obviously, in practice most pairwise comparisons are never made. Instead the comparisons that are considered are restricted to much smaller subgroups of compounds. For a subgroup 1/Kth the size of N, there are roughly (N/K)
2
/2, pairwise comparisons per group. Thus, if N is between 1000 and 10,000, and the subgroup size K is 1/100 the size of N, there may be between roughly 50 and 5,000 pairwise comparisons. With such efficiency gains in subgrouping, there is a compelling interest in a flexible and fast way of forming and organizing subgroups. Such a flexible and fast technique is provided by using identified cyclic systems.
A cyclic system browsing index partitions a large compound collection into interesting and non-overlapping subgroups, and thereby, enables a user to realize the preceding efficiencies in constructing useful pairwise comparisons. Constructing a comparable number of subgroups using conventional substructure and similarity searching methods is a time consuming and error prone operation.
As wil
Johnson Mark A.
Xu Yong-jin
Baran Mary Catherine
Hoff Marc S.
Pharmacia & Upjohn Company
Pharmacia & Upjohn Company
Wootton Thomas A.
LandOfFree
Chemical structure identification does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Chemical structure identification, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Chemical structure identification will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3322976