Stephen R. Heller
US Department of Agriculture
Agricultural Research Service
BARC-W, Bldg. 011A, Room 164
Beltsville, MD 20705-2350 USA
Phone: 1-301-344-1709, FAX: 1-301-344-1823
Telex: 258594 MDCL.UR

Keywords: User Interface, Chemical Structure Searching, Offline Query Formulation, PC Software-Host Connection

Abstract: A brief overview of the reasons for needing PC based user interface software for offline chemical structure query formulation are described.


The major problem facing users of computer systems today is the difficulty of use of both online and PC based software and databases. The lack of good user interface management systems (UMIS) has stalled the extensive usage of many computer systems. The protocols to access online systems are sometimes difficult to learn. Once connected to the computer system knowing the commands and responses needed to use a search system, either online, in-house, or PC-based is not a simple matter. Many systems use odd combinations of keystrokes, special command words, or difficult input strings. Of particular interest to those in the online field who use databases in the field of chemistry are the new software packages designed to provide easier access to structure searching systems. This brief overview of the situation is designed to provide a introduction to the papers in the chemistry session of this year's online conference.


The presence of different online host systems with both bibliographic and structural databases has left many users to learn the intricacies of a particular host logon protocols and language. A number of software packages have been developed to assist users in their online searching of bibliographic databases. The Sci-Mate series of software from ISI, Pro-Cite, and In-Search, to name but a few, are designed to help one quickly and efficiently perform a search for the relevant literature citations of interest.

With the increased usage of chemical structure search systems the problem of learning how to search for a compound or class of compounds has increased considerably. Chemical structure searching is a very complex process. Different chemical tautomer structures, variable bonds, salts, and variable substituents at variable positions, are among the issues one is confronted with when considering how to do a chemical structure search. The fact that different software systems (CAS ONLINE, DARC, SANSS, MACSS, HTSS, and the soon to be released Softron structure search system) all have slightly different conventions, as well as different capabilities, makes it difficult for the novice and the experienced user to perform the actual search they want. For the inexperienced or infrequent user it is even more of a problem. While more chemical information is being made available in computer readable form, it is not necessarily easier to use. This situation is not unique to either chemical information, in-house databases, or online databases. In the entire field of computers, the use of software and programs is getting more complicated as time goes on. Only the PacMan program didn't need a users manual. Today, even with a users manual, a training course, and a hot-line phone to help answer questions, easy and proper usage of most systems is lacking. The Macintosh window system has become a defacto industry standard. The move towards easier to use interfaces, such as the Windows software for IBM-PC's to look like the friendlier Macintosh interface, is becoming the most important issue in the computer world. Faster computers, more storage, color graphics, and other features cannot be effective if the user can't easily use the software system.


Vendors have a goal of wanting to increase usage of online systems. To have more users and more easy to generate accurate queries, a number of vendors in the past year have come out with software designed to help the user. From PC's and workstations which are used to connect to online systems the first generation of front-end, user friendly software is now available. The papers in the chemistry session will emphasize two areas - chemical structure searching and the chemical workstation. While these software systems have only started moving towards the goal of an easy to use interface, the fact that there are a number of such packages now available bodes well for the user.

In the area of chemical structure searching, the complete formulation of a question can involve a number of issues, some of which are often opaque to the user. Questions such as what elements can be substituted at a given atom site, what types of bonds are acceptable for a given chemical fragment query (single, double, ring, chain, aromatic, and so on), and does the system have explicit or implicit substitution, are just a few of the issues facing a user. Software packages such as STN-Express, MOLKICK, Chem-Connection, and CHEMLINK are primarily structure drawing programs which are designed to make it easier for both the novice and experienced user to quickly formulate a query without the need to roam through the long and detailed manuals of the various structure searching systems.

While the different software programs have their own unique characteristics, a few general remarks are valid for all. Firstly, they are all mouse driven programs. Next, they all allow the user to create or formulate a chemical structure query offline. This saves both telecommunication costs as well as connect time costs. Thirdly, they all allow for downloading of answers or hits, which allows one to browse at a later time without the additional telecommunications and connect time costs. Additionally STN-Express has text searching features which are absent in the other software packages.

Whereas some of these programs work with only one vendor, one package, MOLKICK, works with all three online chemical structure search systems, CAS ONLINE, DARC, and the Softron system. MOLKICK is the only IBM PC program which is memory resident. MOLKICK does not have the communications and graphics software built in as part of the package. Chem-Connection is the only Macintosh based package. STN-Express,developed by CAS for its own STN-International system, is the only package which has a structure drawing capability, offline search strategy formulation, and automatic logon.


As was pointed out by Martha Williams and others, the most recent being the October 1988 issues of Monitor, most of the online usage comes from very few users. The Monitor figures quote Renate Mackay, who says that "The analysis [of the online market in Germany] shows that the largest slice of the income, 60%, is accounted for by only 13% of the user population. When the most active 45% of users are considered, they account for 95% of the income." One reason, besides cost, which keeps users away is the difficulty of using an online system. Clearly these four new programs for offline chemical structure query formulation are the first in a number of steps needed to attract more chemists to these systems. However one must wonder if the vendors and database owners of the chemical structures files and software systems really have long term strategic plans or short sales goals. While experienced and frequent users will gain considerably from even these expensive and rough-cut first releases front-end chemistry software packages, such pricing schemes are not likely to win over many new users. While pricing schemes are not the subject of this presentation, it does seem those selling these software packages have yet to realize their significant income must come from the actual search and display of information in their database. Income from manuals, training, and front-end software, while they don't have to be loss leaders, are not going to generate sufficient revenue to cover database costs. Only database usage will do so. Hopefully the next releases of these user interface management system programs for chemistry will be even more user friendly, powerful and less expensive.