NIH/EPA Chemical Information System
G. W. A. MILNE*
National Institutes of Health, Bethesda, Maryland 20502
S. R. HELLER
Environmental Protection Agency, PM-218, Washington, DC 20460
Received December 13, 1979
A network of over a dozen interactively searchable chemical data bases has been made available
for worldwide general use.
The NIH/EPA Chemical Information System (CIS) was started in 1973 as a joint project in
mass spectrometry1 and structure searching2 between these two agencies and has, over the years,
developed with the additional cooperation of other U.S. Government agencies,3 as well as other
organizations in the United States and elsewhere. A preliminary report of the prototype system
was described over 3 years ago.4
Since that time, nine new components have been added to the Chemical Information System
(CIS), regular updating schedules have been established for all CIS components, and the
problems of linking between different data bases have been solved. As a result, the CIS is now
best viewed as a network of chemical data bases, and it is the purpose of this paper to describe
the function and utility of this system.
Current CIS Components (Spring 1980)
OPERATIONAL:
SANSS -Structure and Nomenclature Search System ($60/hr)
MSSS -Mass Spectral Search System ($36/hr)
CRYST -X-ray Crystallographic Search System ($36/hr)
CNMR -Carbon 13 NMR Search System ($36/hr)
MLAB -Mathematical Modeling System ($60/hr)
RTECS -Registry of Toxic Effects of Chemical Substances ($36/hr)
CAMSEQ-11 -Conformational Analysis Programs ($60/hr)
OHMTADS -Oil and Hazardous Materials Assistance Data Base ($36/hr)
PDSM -JCPDS Powder Diffraction Search Match ($60/hr)
FRSS -Federal Register Search System ($60/hr)
XTAL -Single Crystal Reduction and Search System ($36/hr)
WDROP -Water Distribution Register of Organic Pollutants ($36/hr)
NMRLIT -NMR Literature Search System ($36/hr)
UNDER DEVELOPMENT:
IR -Infrared Spectral Search System
CTCP -Clinical Toxicity of Commercial Products
CONGEN -Structure Generation
NUCSEQ -Nucleic Acid Sequences
Figure 1. Schematic diagram of the NIH/EPA Chemical Information System.
DESIGN OF THE CHEMICAL INFORMATION
SYSTEM
A schematic representation of the CIS is shown in Figure 1. Each of the "peripheral"
components consists of a data base, together with the programs for interactive searching through
that data base. Identification of compounds, all of which have CAS Registry numbers, from data
is possible in these peripherals.
If it is desired, on the other hand, to search for any substances with a specific structure or
substructure, this can be accomplished in the central component, the Structure and Nomenclature
Search System (SANSS). The entire CIS is searched by SANSS in a single operation, which
retrieves the Chemical Abstracts Service (CAS) Registry number for all compounds that respond
to searches. Further, SANSS informs the user as to which CIS components contain data on these
compounds. The user is then free to transfer to any component where, using the CAS Registry
number, the appropriate data can be retrieved and displayed.
This design has evolved in large part in response to pressures from users of the system. Its
strength is that it permits one to use the system in several quite distinct ways, and it therefore is
used by different people with different motives. As an example, the chemist studying water
pollutants use MSSS to identify a chemical whose mass spectrum has been measured. Once the
substance is identified, in terms of its CAS Registry number, information concerning its toxicity
is retrieved from the National Institute of Occupational Safety and Health (NIOSH) Registry of
Toxic Effects of Chemical Substances (RTECS), or spill response information is obtained from
the EPA Oil and Hazardous Material Technical Assistance Data System (OHM-TADS). These
are all CIS components, from which data retrieval is a routine procedure.
Alternatively, an administrator, requiring specific information concerning a particular
chemical substance, can locate that substance in SANSS, using either a systematic name or
common synonym. Then the Registry number may be used to retrieve, for example, recent
citations in the Federal Register for the chemical in question.
The use of unique Registry numbers to identify chemical substances leads to a simple and
unambiguous system, and because of this, a decision to use the Registry number as the ultimate
identifier for chemical substances within the CIS was made at an early stage of its development.
A similar policy is now in effect within the EPA5 and has been recommended as a general
government policy by the Council on Environmental Quality.6 Other systems also use this
identifier, and it has thus become the link between various systems as well as between the
different components of the CIS. Such s link can be exploited in a variety of ways; as an
example, the Registry numbers retrieved by a substructure search in SANSS can be transferred to
the Lockheed Corporation's Dialog system.7 Here, substructure searching is based upon
nomenclature and thus less precise perhaps than SANSS, but all the CAS literature citations to a
compound can be easily retrieved by using its Registry number. By use of the Registry number
as a link, the substructural searching power of the CIS and the enormous bibliographic scope of
Dialog can be used synergistically.
NEW COMPONENTS OF THE CIS SINCE 1977
(1) Registry of Toxic Effects of Chemical Substances (RTECS). Maintained by NIOSH,
this file currently contains over 70,000 toxicity measurements on some 41,000 compounds.
Within the CIS, it can be searched by animal species, dosage route, and toxicity level.8 Like all
CIS files, it can be searched by chemical structure in SANSS.
(2) Oil and Hazardous Materials-Technical Assistance Data System (OHM-TADS). The
Office of Water and Waste Management of EPA maintains this data base of - 1000 major
chemical substances. Over 120 different types of information ranging from toxicity data to the
recommended response to spills of the substance, are available on each compound.
(3) Powder Diffraction Search-Match System (PDSM). The Joint Committee for Powder
Diffraction Standards-International Data Centre has assembled a file, currently of 33591 X-ray
powder diffraction patterns, which is very useful for the identification of both organic and
inorganic materials. This data base is searchable in the PSDM component of the CIS.9 Since
powder diffraction measurements are most commonly made upon mixtures of phases, PDSM
uses a reverse search procedure, which checks to see if each library pattern can be "contained"
within the data derived from the unknown. As a library pattern is identified within the input
data, that entire pattern can be subtracted from the data and the whole search repeated upon the
residue. In this way, a qualitative and semiquantitative analysis of mixtures can be
accomplished.
(4) Federal Register System (FRSS). One of the critical areas of chemical information for
use by regulatory agencies and commercial organizations are the chemicals cited in the U.S.
Federal Register. Thus, to provide a total picture of chemical identification, chemical property
(e.g., toxicity), and chemical regulations, the CIS has developed a data base and search system
for the U.S. Federal Register.10 The Federal Register Search System (FRSS) contains citations to
chemicals in the Federal Register from Jan 1, 1978, to the present. The file is updated weekly
and, in mid-1980, contained over 55000 citations to chemicals found in the U.S. Federal
Register. The system is searchable by chemical name, CAS Registry number, agency, type of
regulation, and so forth.
(5) X-ray Single Crystal Search System (XTAL). The X-ray single crystal system is
derived from the NBS/JCPDS publication titled "Crystal Data Determinative Tables." The data
base will contain over 50,000 entries of crystal studies done which include information on the
space group, density, and unit cell. The data base is expected to be updated annually.
(6) WaterDROP (WDROP). The WaterDROP system is a data base developed by the EPA
laboratory in Athens, GA.11 The data base is a collection of semievaluated references and
citations to chemicals found in water systems in the United States and elsewhere. Hence, it
serves as a Distribution Register of Organic Pollutants (DROP) found in water. At present, the
system contains 10,600 citations to chemicals found in all types of water, including drinking
water.
(7) Clinical Toxicology of Commercial Products (CTCP). The Clinical Toxicology of
Commercial Products (CTCP), now in its fourth edition,12 contains toxicity information for over
3000 chemicals which are part of formulations of some 20,000 products. The system, currently
being tested and expected to be available for operational use in the CIS by the end of 1980, is
searchable in a number of ways. The searchable parameters include chemical name (s), CAS
Registry number, manufacturer, toxicity rating, and so on.
(8) Data Base Independent CIS Components. Two CIS components are processors rather
than data-retrieval programs. The Mathematical Modeling Laboratory (MLAB), developed at
NIH,13 is a powerful general-purpose statistical analysis program. CHEMLAB (formerly called
CAMSEQ-II14), which permits conformational analysis of user-defined structures, is also a CIS
component.
(9) Structure and Nomenclature Search System (SANSS). Work was first started on a
substructure search system (SSS) for the CIS some 10 years ago.15 Since then the software has
been continually expanded and the name of the system was changed from SSS to SANSS
(Structure and Nomenclature Search System) so that the full capabilities of the system were
properly recognized.2
Every chemical substance in the CIS has been unequivocally identified with its Chemical
Abstracts Registry number much like Social Security numbers are associated, hopefully
uniquely, with individuals. This process has also been completed for a considerable number of
chemicals that are not in any CIS data base but which comprise integral files. Examples of this
are the five substances that are regulated by Section 311 of the Clean Air Act or the 8981
chemicals which appear in the Merck Index.16
Each of the - 190,000 Registry numbers that has been so obtained was used to retrieve from
the CAS Registry the nomenclature and structure records for the compound. These records
constitute the SANSS data base.17
Perhaps the simplest method of searching within SANSS is with a compound name. a full
name (2-chlorophenol) or a partial name (chlorophenol) may be used, the latter of course giving
more hits. Any name may be truncated on the left (:orophenol, which will retrieve chloro-and
fluorophenols) or on the right (chlorophen:, which will retrieve chlorophenanthrenes, for
example, in addition to chlorophenol).
a higher degree of precision can usually be obtained in SANSS with a structural search. The
"query structure" used in such searches must be generated by the user, with the help of the
structure building commands in SANSS. Once the query structure is defined, it can be used in
searches for all compounds containing a fragment centered upon a particular atom, a ring, or
system of rings or the entire query structure either imbedded in a larger structure or as an exact
match to the file structure.
Finally, compounds containing specific functional groups can be retrieved by using specific
(CIDS) keys for those groups, 18 searches for molecular weight and molecular formula are also
permitted.
When a SANSS search of any sort is completed, the number of hits is reported to the user and
the appropriate Registry numbers are stored in a temporary file, which can be used in subsequent
Boolean operations or to display the structure and properties of the compounds that were
retrieved by the search. If such a display is requested, up to five items of information concerning
each compound may be printed at the terminal. These are the Registry number, the CIS data
bases and non-CIS data bases that contain information pertaining to the compound and the
compound's molecular formula, structural formula, and all known names and synonyms. An
entire file of Registry numbers may be used by another CIS component, such as MSSS, to
retrieve the mass spectra or in a non-CIS search system, such a Lockheed's Dialog, 7 to retrieve
the corresponding literature citations.
EXAMPLES OF THE USE OF THE CIS
In this section, three examples are given in which the CIS was used to help solve specific
problems of different sorts. The interfile linking in the system is, as will be seen, very useful in
data-retrieval problems such as these.
In the first case, which is shown in Figure 2, a water pollutant has been found to exhibit in its
mass spectrum intense peaks in the m/z ranges 235-237 and 293-295. The base peak (100%
intensity) in the spectrum is at m/z 237, and when this is entered into the PEAK search options of
MSSS, 19 a total of only 23 spectra in the data base of 32191 are found to have their base peak at
this m/z value. A second peak, at m/z 295, reduces the number of hits to two and the third peak,
at m/z 293, limits the number of retrievals to just one. The pollutant is thus tentatively identified
as tetraethyllead, CAS Registry No. 78-00-2. In search of confirmatory data, the carbon-13
NMR spectrum is next retrieved. This is accomplished by the command CSHOW, transfers the
user from MSSS to CNMR, and then uses the Registry number to identify the correct spectrum,
which is printed out.
The identification now considered fairly certain, the next task is to assess the risk implied by
the discovery of this substance in a water supply. As shown in Figure 3, the command
"TSHOW1" leads to transfer to the RTECS component. Here, lookup, retrieval, and printing of
all the toxicity data available for the compound in file 1 take place, and then the user is returned
to the DNMR component.
As can be seen, the substance is particularly toxic in most animals species, and so, in order to
learn more about the risks posed by this substance, a transfer is requested (GO OHM-TADS) to
the Oil and Hazardous Materials file. Once there, the information fields pertaining to production
sites (prd), degree of hazard to public health (hel), and recommended drinking water limits (drk)
are sought, and the command OTSHOW 2/4 results in the retrieval and listing of all information
in those fields for tetraethyllead.
From the information obtained in this session, it has become clear very quickly that this is a
serious pollutant and steps to deal with this situation should be taken immediately. From initial
access to the CIS to completion of the session required - 4 min, and the cost of the session, as
reported to the user by the COST command, was approximately $3.20.
In a quite different use of the system, it can provide a rapid means of retrieval of literature
citations to specific structures or groups of structures. In the example shown in Figure 4, the
problem was to locate all papers published since 1997 on compounds having the basic ring
skeleton of colchicine. In SANSS, the command NUC 67u7 creates this ring system, and the
option RPROBE is used to search for all compounds containing it. Any substitutes are allowed
and the result is 25 compound, whose Registry numbers are stored in file 1. Four names and
synonyms of a few of these are listed and can be seen to include colchicine itself, colchicine,
colchicoside, and the commercial product N-deacetyl-N-methyl colchicine. The entire file of 25
Registry numbers is next stored in the bubble memory of the Texas Instruments' Model 765 that
is being used as a terminal. This is done with the command "Dialog 1," which locates the file of
numbers and holds them in memory in a format appropriate for use in the Dialog system. When
the storage has been completed, the user transfer through the communications network from the
CIS host computer to the Lockheed Dialog host, whose network address is 41520. When a
connection is made here, as is shown in Figure 5, Dialog's file No. 4 (Chemical Abstracts, 1977-1979) is accessed, and the Registry numbers are unloaded, five at a time, from the terminal's
TYPE PEAK, MIN INT, MAX INT
CR TO EXIT, 1 FOR REGN, QI, MW, MF AND NAME
USER: 237, 100, 100
FILE # # REFS M/Z PEAKS
1 26 237
NEXT REQUEST: 293, 30, 40
FILE # #REFS M/Z PEAKS
2 1 237 293
NEXT REQUEST: 1
REGN Q1 MW MF NAME
78-00-2 704 324 C8H20Pb
Plumbane, tetraethyl - (8C19C1)
Lead tetraethyl
Lead tetraethyl-
Tetraethyllead
Tetraethylplumbane
Option? CSHOW 2
CAS RN: 78002
C9
*
*
C5
*
*
C7*C3*PB*C2*C6
*
*
C4
*
*
C8
Plumbane, tetraethyl-
WLN: 2 4-PB-
CAS RN = 78-00-2 CMR# 601 MW = 323.35
C8 H20 PB1
EIDGENOSSICHE TECHNISCHE HOCHSCHULE
Solvent: UNKNOWN
SHIFT MULT INTENS ASSIGN
15.6 T 80 2
10.6 Q 100 6
Figure 2. Identification by mass spectrometry and carbon-13 NMR spectroscopy of an organic pollutant.
memory. In this way, a total of 470 citations to the 25 compounds are located, and one citation is
printed out in the figure, as an example.
In a real case, of course, a sensible next step would be for the user to reduce this list of 470
citations to those of particular interest at the time. This is easily done by, for example, a subject
search in Dialog, followed by the appropriate Boolean combination of files.
Figure 3. Retrieval of acute toxicity and spills response information for tetraenthyllead.
All of this searching and retrieval can be done in under 15 minutes, including the transfer
from the CIS to Dialog. The cost within the CIS was $3.47 and the Dialog charge was $1.86 for
a total cost of $4.33.
There are no other practical approaches to this sort of a problem, because the structure search
in Figure 4 cannot be done manually, and attempts to conduct a search of this sort with
compound names will rarely be exhaustive, as a consequence of the inadequacies of chemical
nomenclature. The cost in time and money therefore can only be compared to the alternative of a
skilled scientist, browsing in the library for many hours; neither the cost of the scientist's time
nor that of maintaining the library is inconsiderable, and in this sort of a comparison, the online
retrieval system, such as the CIS and Dialog, are very competitive.
Figure 4. Retrieval and storage of CAS Registry numbers for all compounds with a specific ring
skeleton.
A much more detailed structural search in the CIS is illustrated in a third example, shown in
Figure 6, in which information is sought concerning the expected NMR chemical shift of a
methylene carbon substituted by a bromine atom. The necessary substructure search is carried
out in SANSS, where searching is first limited, by using the SELECT command, to the file
corresponding to the CNMR data base. Then a query structure, -C-C-Brm us constructed and an
FPROBE search is carried out for all compounds containing a C-Br residue. The TERMA
command limits to two the number of nonhydrogen neighbors permitted for atom 2 and thus
defines the substructure as -C-CH2-Br. Then the file of 110 bromo compounds is inspected with
the SUBNMR search option for all compounds containing this substructure.
A total of 42 compounds are retrieved in this search, and their Registry numbers are stored in
temporary file 2. The user now requests transfer to the CNMR Search System, and , once there,
issues the command SUB, as shown in Figure 7. Atom 2, the carbon in the in the query structure
bearing the bromine, is nominated, and the program uses the 42 Registry numbers in file 2,
together with the corresponding CNMR data, to find that a shift for such a carbon has been
assigned and reported 22 times. Five of the reported shifts are diagnosed as outliers, and from
the remaining 17, the average shift is calculated as 32.2 ppm, with a standard deviation of
3.2 ppm.
Figure 5. Transfer of a file of CAS Registry numbers to Dialog and retrieval of Chemical
Abstracts citations.
The cost of this complete inquiry is $3.79, and it was completed in a little more than 5
minutes. Again, this is a search which is almost impossible manually; the compilation of tables
of CNMR shifts, which contain items such as this value, has required many expert
spectroscopists to spend considerable time, and while such tables are very valuable, they do no
cater to the chemist who is interested in the likely chemical shifts for the carbons of a particular
unusual structure. This SANSS/CNMR linked search can tolerate any structure ; its limitation is
that the CNMR data base, for reasons discussed in the next section, is not large enough to
support wide-ranging queries.
Figure 6. Substructure search in the CNMR data base for all compounds containing the bromo
ethyl moiety.
Figure 7. Retrieval and statistical analysis of all carbon-13 chemical shifts for the bromo ethyl
moiety.
MANAGEMENT ASPECTS OF THE CIS
Development and operation of the publicly available CIS have been in progress for about
eight years now, and during that time, it has become increasingly clear where the major
difficulties lie. This section contains a discussion of some of these points, consideration of
which is relevant to continuing development of the system.
(1) Programming and Data Acquisition. In the early days of the project, the cost of
software development was considerable, but as generalized software is being implemented, the
proportion of total development costs that support programming is steadily diminishing. By
contrast the cost of data collection and evaluation has emerged as the largest continuing cost
within CIS development and also one of the most intractable problem areas, because it involves
both data acquisition and evaluation, both of which present special difficulties.
Much of the early collection of data for the CIS was done on an ad hoc basis in that mass
spectroscopists, for example, were contacted and asked to provide spectra for the data base.
Most of the known collections of spectra have been acquired and merged into the growing file
and, consequently, this approach has in recent years become less fruitful. It also has the added
disadvantage that it allows little control over the content of the file. For these reasons, mass
spectra are now being acquired by direct measurement on those compounds whose presence in
the data base is felt to be important. This approach permits some control of the file content and
quality but is very costly; acquisition and entry of a mass spectrum into the file in this way can
cost as much as $250.
The primary literature provides very little assistance in problems of this sort. This is
generally because the cost of page space in now so high that most journals refrain from
publication of actual data. Thus, those interested in collection of data are in the tantalizing
position of knowing that data exist but are directly available. This is a gloomy situation, relieved
only by the light shed upon it by a program announced in 1977 20 by the Chemical Society in
London. Those wishing to publish X-ray crystallographic papers in the journals of the Chemical
Society are now required to submit all numeric data together with the manuscript. The data may
be reviewed by referees but are generally not published in the journal; instead, they are sent to
the Crystal Data Centre at Cambridge University where they are used in the compilation of the
Cambridge Crystal file, an authoritative data base, which is leased from the Data Centre by most
countries in the Western world. The income from these leases supports the Data Centre, and the
cost of acquiring the raw data is reduced to almost zero. This experiment appears to have been
quite successful, but it has yet to be emulated by other chemical societies.
The field of data evaluation is very poorly developed, and there is general agreement only
that this is an important activity if data bases are to be reliable. The cost of data evaluation is,
however, considerable, and in many cases, it is not even clear how data an be evaluated, at any
cost, without actually repeating the measurements in question. There have been some tentative
efforts directed at validation of data in the CIS files, and the experience of the National Bureau of
Standards has proved to be very helpful in this area. It remains a fact, however, that most of the
data in the CIS is unevaluated, and how it might be evaluated economically is not obvious.
(2) User Support and Marketing. The CIS currently has more than 900 users from over
400 user organizations who carry out between 30,000 and 40,000 transactions per month.
Consequently, it is important that there be adequate communication between them and the
system managers, and a number of approaches have been adopted to ensure that this is the case.21
A monthly CIS newsletter is published and sent to over 2000 individuals. This publication,
which is typically about four to six pages in length, provides general information about the CIS.
Those who use the system have access to an online news system, which provides headlines at log
in and the full news message upon request. This news system is the best means of ensuring the
users are apprised in detail of changes in the CIS.
As far as communication from users and others is concerned, two procedures have proved
effective. A toll-free telephone is maintained on a 24-hour basis and is used perhaps 12 times per
day, mostly by nonusers of the CIS, for inquiries of various sorts. Users of the system can
submit comments or complaints using a CIS option, called COMMENT. Such comments are
collected and reviewed daily. On average, about one comment per day is received and the
appropriate response is provided as soon as is possible, typically within 3 days.
(3) Future Directions. One of the major administrative concerns with the CIS has been its
cost and implicit open ended nature. The CIS was designed to be a public system, and it was
decided that an annual $300 subscription fee for access to the system should be levied. (This fee
has been waived for educational institutions as of May 1980.) Use of a commercial computer is
consistent with OMB Circular A7622 and leads to the generation of funds which is used to defray
the cost to the government of the user support discussed in the preceding section. Current
projections suggest that the subscription fee and other revenues will match the support and
marketing costs in 1981. After this time, these revenues will be used to defray the costs of
maintenance of the CIS. In this way, provided use of the CIS does not decrease (which is hardly
likely given the recent doubling of growth per year for the last two years), it is possible that it
could become self supporting at the operational level by 1982.
As these efforts to control costs are continuing, some attempts are being made to control the
directions in which the CIS is growing. A Steering Committee now reviews questions such as
addition of new components and modification of existing modules, and, as this Committee is
open to all the many groups, governmental and otherwise, which have collaborated in the
development of the system, it is hoped that it will function democratically in guiding the future
development of the CIS. Lastly, a Science Advisory Board, comprised of scientists from
industry and academia, has been established to assure that the highest possible scientific levels
are maintained in the system.
Thus, the CIS has evolved to the point where a solid system has been built and is being used.
The tasks ahead now are to develop a smoothly functional maintenance and operation staff so
that the main government efforts can be directed to further developments in areas deemed
relevant to those senior-level decision-making managers who control the funding of such projects
as the CIS.
REFERENCES AND NOTES
(1) S. R. Heller, Anal. Chem., 44, 1951 (1972); S. R. Heller, H. M. Fales, and G. W. A. Milne, Org. Mass
Spectrum., 7, 107 (1973); S. R. Heller, H. M. Fales, and G. W. A. Milne, J. Chem. Educ., 49, 725 (1973); S.
R. Heller, D. A. Koniver, H. M. Fales, and G. W. A. Milne, J. Chem, Doc., 13, 130 (1973); S. R. Heller, H.
M. Fales, G. W. A. Milne, R. J. Feldmann, N. R. Daly, D. C. Maxwell, and A. McCormick, Adv. Mass
Spectrum., 6, 1037 (1975); S. R. Heller, G. W. A. Milne, R. J. Feldmann, and R. S. Heller, J. Chem. Inf.
Comput. Sci., 16, 176 (1976); G. W. A. Milne and S. R. Heller, Am. Lab. (Fairfield, Conn.), 8, 43 (1976); S.
R. Heller, R. S. Heller, A. McCormick, D. C. Maxwell, and G. W. A. Milne, Adv. Mass Spectrum., 7B, 985
(1977).
(2) R. J. Feldmann, G. W. A. Milne, S. R. Heller, A. Fein, J. A. Miller, and B. Koch, J. Chem. Inf. Comput. Sci.,
17, 157 (1977).
(3) These Agencies were the National Institutes of Health, the Environmental Protection Agency, the National
Bureau of Standards, the National Institute of Occupational Safety and Health, and the Food and Drug
Administration. In addition support from the U.K. Department of Industry has been made available since
1971.
(4) S. R. Heller, G. W. A. Milne, and R. J. Feldmann, Science, 195, 253 (1977).
(5) EPA Internal Order 2800.2 states that "Any computer-based Agency...system...containing data...on..chemical
substances shall contain the CAS Registry number for each chemical substance..." See ref 6, p 38.
(6) "The Feasibility of a Standard Chemical Classification System and a Standard Chemical Substance
Information System," report to the Congress from the Council on Environmental Quality, Stock No. 041-011-00039-4, U.S. Government Printing Office, Washington, D.C., July 1978.
(7) The Lockheed Corporation serves as a vendor for the CAS Condensate and CASIA data bases. Dialog, a
Trademark registered with the U.S. Patent and Trademark Office, is a proprietary product of Lockheed
Information Systems Inc., Palo Alto, CA 94304.
(8) J. R. McGill, S. R. Heller, and G. W. A. Milne, J. Env. Path. and Tox., 2, 539, (1978).
(9) R. G. Marquart, I. Katsnelson, G. W. A. Milne, S. R. Heller, G. G. Johnson, and R. Jenkins, J. Appl.
Crystallogr., 12, 629 (1979).
(10) G. Marquart, L. Marquart, S. Mintz, J. McGill, J. McDaniel, S. R. Heller, and G. W. A. Milne, Online
(Weston, Conn.), 4, 45 (1989).
(11) A. Alford, R. Potenzone, S. R. Heller, S. Mintz, and G. W. A. Milne, publication in preparation.
(12) R. E. Gosselin, H. C. Hodge, R. P. Smith, and M. N. Gleason, "Clinical Toxicology of Commercial
Products," 4th ed., Williams and Wilkins, Baltimore, MD (1976).
(13) G. D. Knott and R. I. Shrager, Assoc. Comput. Mach., SIGGRAPH Not. 6, 138 (1972).
(14) R. Potenzone, Ph.D. Thesis, Case Western Reserve University, February 1979. See also R. Potenzone and A.
J. Hopfinger, "Structural Correlates of Carcinogenesis and Mutagenesis. A Guide to Testing Priorities,"
Proceeding of the 2nd FDA Office of Science Summer Symposium, August 28, 1978.
(15) R. J. Feldmann and S. R. Heller, J. Chem. Doc., 12, 48 (1972).
(16) "The Merck Index," 9th ed. Merck & Co., Inc., Rahway, NJ 07065, 1977.
(17) G. W. A. Milne, S. R. Heller, A. E. Fein, E. F. Frees, R. G. Marquart, J. A. McGill, and J. A. Miller,
Computer Science, 18, 181 (1978).
(18) "Handbook of CIDS Chemical Search Keys," Fein-Marquart Associates, Inc., Baltimore, MD 21212,
November 1973.
(19) S. R. Heller, Anal. Chem., 44, 1951 (1972).
(20) See the Notice to Authors, J. Chem. Soc., Chem. Commun., No. 3 (1977), No. 3 (1978).
(21) For details on the CIS operations and availability of the CIS, please contact Kay Pool, CIS Project,
Information Science Corporation, 2135 Wisconsin Avenue, NW, Washington, DC 20007 [(202) 298-6200 or
(800) 424-2722].
(22) "Circular A76" (Office of Management and Budget, Washington, DC, August 1967) states, in essence, that
all attempts should be made to ensure that systems such as the CIS, operate in the private sector.
