Reverse search

The large peak at scan 502 (Fig. 1.7) does not interfere with the ability of the software to quantify the sample. Although the compound eluting at scan 502 was not one of the target compounds in the library being reverse-searched, it was possible to identify it by forward-searching the NBS library present on the system. The greatest similarity was in the comparison of the unknown with the spectrum of benzaldehyde. 

This TEXT is used in the various explanation and tracing facilities. Whenever the inference engine reaches one of these FACTS it either continues its search, if possible, or proceeds another level deeper in the reverse search and tries to prove that FACT. 

The various match factors calculated by the matching program are listed in Table I. The overall match factor (PT) is a combination of forward and reverse searching techniques. It takes into account the deviations in intensity of the sample spectrum peaks with respect to the candidate spectrum peaks and vice versa for all peaks in both spectra. The pattern correspondence match factor (PC) is a forward searching match factor which takes into account the intensity deviations of sample spectrum peaks with respect to the candidate spectrum peaks for peaks common to both spectra. This factor detects structural similarities, such as substructures, based on common spectral patterns. NC, NS, and NR give an indication of the number of peaks upon which the match was based and in which direction it was most successful. IS and IR indicate the magnitude of the ion current unmatched in each direction. These match factors are similar to those proposed by Damen, Henneberg, and Wiemann (9). 

The data base of some 27,000 powder diffraction patterns that is used in the CIS (5) is in fact a direct descendant of that with which Hanawalt carried out his pioneering work. A problem that arises in connection with this particular component stems from the fact that powders, as opposed to crystals, are frequently impure and so the patterns that are obtained experimentally are often combinations of one or more file entries. A reverse searching program, that examines the experimental data to see if each entry from the file is contained in it, has been written after the general approach of Abramson (23), and seems to cope with this particular difficulty. It is currently running in test on the NIH PDP-10 and will be made available to the scientific community during the latter part of 1978. 

The starting point of a reverse search is often from a lead product that is currently on the market but in need of improvements, with a given set of properties that should be modified. The search questions include ... 

The reverse search is the most often encountered search question in product engineering, but there are few organized and convenient search engines. 

Thus, we execute three forward searches plus interpolation, which has the effect of a reverse search. 

Other important handbooks include Chemical Properties Handbook (Yaws 1999) and the CRC Handbook of Chemistry and Physics (Weast and Tide 1989) Tables of Physical and Chemical Constants (Kaye and Laby 1986) is a more compact handbook of physical and chemical data. One should be on guard that sometimes, when experimental results are not available, the editors may list estimated values in a handbook, which are of less certain accuracy. A printed handbook normally has only limited reverse search capability, of going from a set of properties to the structures that have these properties. 

A simple spreadsheet, such as Microsoft Excel, can serve as the foundation of a database that has forward and reverse search capabilities. For instance, a table of normal alkanes, together with their densities, boiling points, and melting points, can serve as the starting point. If we want to know all the normal paraffins that boil between 0 and 40 C, all we have to do is to do a sort operation on the boiling-point column and obtain the result that the only paraffin that is in the range is normal heptane with a boiling point of 36.1 °C. For the more advanced Boolean search of normal alkanes that boil between 0 and 40 °C AND melt between —40 and 0 °C, it would be a far more laborious task in a spreadsheet. 

One can also do a reverse search, such as for all compounds that boil between -40 and 0 °C, to speed up the search a great deal. Each of these variables can be... 

It has extensive capability for forward search and reverse search from properties to substances. The properties available include ... 

Do a reverse search from properties to materials. What are all the hydrocarbons... 

Compact information storage and retrieval. A good correlation leads to compact storage of information in the form of equations and coefficient values, which is more convenient for rapid retrieval than tables of numbers or charts and graphs. This property, which is valuable in a forward search from substance to property, is even more valuable in a reverse search from property to substances. 

Often, the correlation is not good, and we need to search for the hidden variable that we have not yet discovered. But when we find a good correlation, it could prove useful in the reverse search for other untried compounds that may have higher or lower camphor smell, even if we do not understand the mechanism of how it works. There is always the hope that, if we know which parameters are important to smell, we may generate one or more hypotheses on the nature of camphor smell this would be followed by predictions and experiments that could lead us to future understanding. 

The reverse search starts from a set of desired properties and asks for substances that possess them. Theoretical knowledge and past experience should be relied upon to suggest where to look, since it is the fastest and least expensive approach. When theoretical knowledge and past experience have been exhausted, then random searches may be the only way to make progress, if the problem is sufficiently important and there is enough budget and patience. Table 7.1 compares some of the requirements and the pros and cons of the guided search and the random search... 

The most important tool in the arsenal of the product innovators is the ability to make predictions on which structure would lead to what properties, as well as what structure modifications would lead to what property modifications. The reverse research from a given set of properties to material that has these properties is even more important in creating new products and in modifying existing products. In most cases of molecular properties, it is more realistic to depend on empirical correlations between structure and properties. Forward and reverse searches are currently only available for simple physical-chemical properties, such as boiling points and densities such a facility is still not available for biological properties, such as narcotic and antibiotic activities. The development of such search engines would have a tremendous impact on the productivity of product innovators. 

F. E. McLafferty. This algorithm, called Probability Based Matching or PBM, employs a large number of fragment ions in a reverse search mode (4) to provide reliable identification and quantitation of an individual, specific compound even though the mass spectral measurements are derived from a mixture of two or more compounds (5). 

In other words, we could pull out ail of the computer lines with property "A", could see what kinds of properties, B" were associated with property "A", and then run the reverse search for property "B" to see whether we had missed any previous "k s" or not. In this way we obtained the intersection of two hyperspace axes and therefore a primitive form of clustering in an alphanumeric system. 

Two types of library search have been developed. The first type, called forward search, compares the new spectrum with the spectra stored in the library and looks for the best match of the spectra. The second type, called reverse search, checks for the possible presence in the new spectra of a spectrum chosen in the library. 

In the forward search, all of die ions in both the library spectrum and the unknown specdiun are used in the calculation of the fit factor, while in the reverse search only the ions in the library specdiun are used in this calculation. Forward searches work well only for pure compounds, while die reverse search is admirably suited for dealing widi die composite spectra fiom mixtures. 

No MDA was detected in the water or 3% acetic acid extract solutions from either Sample 1 or Sample 2 in the three-day samples using reverse searching (looking for ions at m/z 198, 106 and 182). However, two species could be identified in the extracts. 

The reverse search refers to the fact that the algorithm checks whether a peak from the reference spectrum is present in the unknown (and in the appropriate abundance) and not the other way around. In this way, the reverse search ignores peaks in the unknown that are not present in the reference. The mass library searches are, in fact, much more elaborate, and provide at the end a list of possible matches and for each of them a calculated percentage match. 

The report consists of three parts, (i) a reconstracted ion chromatogram, (ii) a report with name and amount of compounds arranged according to chemical formula, (iii) a reversed search status report containing the data needed to evaluate the results, i.e. found retention time versus theoretic retention time and library matching. 

When peaks are incompletely separated identification may still be possible using a reverse search. The ability of an algorithm to match two or more components in the mass spectrum of a mixture is aided by requiring only that the peaks of the reference spectrum are present in the unknown spectrum rather than the other way round, as for a normal (or forward) search. The hit list of retrieved library spectra should then represent the compounds in the spectrum of the mixture provided that their spectra are present in the reference library. Subtracting the best-hit library spectrum from the mixture spectrum produces a residual spectrum that can then be matched against the other spectra in the hit list in a forward search. In a sequential process identification of the component spectra may be achieved. 

The forward search is the most rapid but demands unknown spectra of pure compounds to produce good results. If however the unknown spectrum includes peaks from unwanted impurities or of a mixture then it will not work correctly. The alternative reverse search strategy although slower is now required, whereby the resulting hit hst shows the best reference spectrum in the library as found in the unknown data. Peaks in the unknown which do not appear in the Hbrary... 

---