SANDRA--Structure and Reference Analyzer Program for the Beilstein Handbook of Organic Chemistry

By Alexander J. Lawson. SpingerVerlag: New York. List price $500.00 (950 DM).

Colleagues here and abroad talk of the bygone days of organic chemistry when everyone knew German and used the Beilstein Handbook of Organic Chemistry the way broken English is the language of chemists and the CRC Handbook is used today. Well there is hope for a comeback for the Beilstein Handbook, the premier source of evaluated data and information on organic chemicals. Besides the 5th edition and all subsequent editions being written in English, the hope lies in the recently released program SANDRA, which is the abbreviation for Structure and Reference Analyzer. The program is very simple to describe, and it takes only one sentence. One draws a structure in an IBM PC using a mouse and hits a key, and the program analyzes the structure and comes up with a set of pointers as to where the chemical would be located in the 340 or so volumes of the Beilstein Handbook.

The Beilstein Handbook of Organic Chemistry, covering the chemical literature from 1830 to 1979 (Basic Series plus Supplemental Series I-V), systematically orders organic compounds by the number of carbon atoms, degree of unsaturation, and substitution patterns. The system is hierarchical and thus produces a unique location for every organic compound. Analogues and homologues are brought together under the system, so such compounds can be easily found. While not a lost art, there are few chemists today who can quickly and easily find a chemical in Beilstein without previously knowing the compound's specific index system number, since a basic knowledge of organic chemistry and German chemical nomenclature are required. (The latter is needed because once you use the molecular formula index, you must be able to know the German nomenclature to pick out your compound from the list of possibilities all having the same molecular formula.) Thus we have a situation tailor made for a computer. This complex but well ordered and structured classification scheme, the Beilstein system, has been programmed into a computer algorithm by Dr. Sandy Lawson, the Managing Director of the 5th edition of the printed Beilstein Handbook.

SANDRA, which is copy protected, comes on one disk and is easily installed on an IBM PC (or equivalent). The program runs best on a hard disk, but it will run on a floppy disk system. The program requires 256K of memory, a Microsoft or equivalent mouse, an IBM EGA graphics board or equivalent (a Hercules board version is due out in late 1987), and DOS 2.0 or later. There is a 78-page manual, which is reasonable, and it has many examples (there are 49 figures), but some commands are not explained (e.g., alternate bond mode), the index is sparse, and there is no detailed example of going from the output of the program to the actual printed Handbook volumes and pages. The program costs $500 (high, but perhaps not so high compared to the cost of the printed volumes, which now run about $20000 per year for the 17 or so Supplemental volumes published each year). Information is available from the publisher regarding volume discounts for multiple copies within an organization. The program, like the Beilstein Handbooks printed volumes, is distributed by Springer-Verlag (ISBN 3-540140046-8).

The program operation is very simple. It takes about half an hour to learn the particular commands of this structure drawing program. (It would be nice if one day there was a standard structure drawing program for all programs for chemists.) The program allows for entering up to 70 non-hydrogen atoms, which seems sufficient for most all purposes. One can use stereochemistry, as well as isotopes, but these have little meaning since these similar compounds will give pointers to the same locations in the Handbook. Once finished the program takes 2-12 s to come up with the set of pointers. The time depends on the complexity of the structure and the type of computer used. I have used both XT and

AT class machines, as well as molecules up to the 70-atom limit. The program has worked every time, and when I have looked up the results in the printed version, the program had indeed pointed me to the right place. Sometimes the pointers are fairly broad (30-40 pages) and sometimes just a few pages. It all depends on the class of compound being analyzed.

The example I have chosen to use in this review is a simple one, 2-pyridinecarboxylic acid (molecular formula C6H5O2N; Chemical Abstracts Service Registry No. 98-98-6). The SANDRA program pointers produced by the program are the following--Haupt-pages, 34 to 51; System-Number, 3249-3251; degree of unsaturation, 2n - 7; Carbon Number, 6; and Supplemental Volume E III/IV 22/1. I went to the library and found the compound starting on the very bottom of page 33 of the Main (Haupt) volume and continuing on through most of page 34. The system number on the page was 3249, and the degree of unsaturation was C,,H2,,702N. I then went to the Supplemental Volume 22 and found additional data on this compound on pages 303-304. On page 303 there were the references to the Beilstein Supplements E I and E 11, where additional data could be found on pages 502 and 30, respectively. SANDRA does not provide the information on these Supplemental volumes. The reason is stated in the manual: "If the required compound was only described before 1930 the entry can always be found using the General Index (H-E 11) (General register). For this reason, SANDRA does not give any specific references to the Series M-E-II in the (pointer) field. However, the volume number (e.g., 22 in this case of E III/IV 22/1) naturally also applies to the single volumes in the same series H to E II." Personally I would prefer the additional information in the pointer field, even if it .may appear to be redundant, since I feel few people will read the manual to find this out.

In addition to this compound, I also entered the 3- and 4-substituted acid, 3-pyridinecarboxylic acid (nicotinic acid) and 4-pyridinecarboxylic acid (isonicotinic acid). The SANDRA output was exactly the same, as would be expected. The actual data for these compounds were readily found in the printed volumes. For nicotinic acid, the data were found on pages 38-39, and for isonicotinic acid, the data were found on pages 45-46. These are both within the range of pages 34-51. Finding the additional data on these two isomers was readily accomplished, as they both follow the first compound by a few pages in each of the Supplementary volumes.

The program should be of great value to working chemists and should be a required part of any organic chemistry course and chemical information course. Hopefully the vendor will find a way to provide copies of the program to universities and colleges for such teaching purposes. My son who is taking organic chemistry in college this year received a short tutorial on Beilstein from a lecturer who had not received the proper training and had made little use of this valuable resource. It was clear from this experience that my son would never consider Beilstein as a source of chemical knowledge. During his spring break I showed him SANDRA, and he said "This I can understand and use".

In summary the program performs well. It does what it says it will do. Its accuracy is good to excellent. Given the wide range of organic compounds, and the varying amounts of information and the differing number of derivatives for a given compound, it would be hard to ask for more from this program. What is needed, and expected to become available within a short period of time, is a template and a handy user guide. After learning how to use the SANDRA, one should have to turn to the manual only for reference. A Macintosh version is not expected soon, as the market for such a version is said to be limited.

Stephen R. Helter, Agriculture Research Service