Page 65 - Visions of the Future Chemistry and Life Science

54 J. M. GOODMAN

there may be billions of possible strategies for making a new molecule.
Each reaction in each strategy must be compared with the literature in
order to discover if similar reactions have been done before, and if the pos-
sible side reactions which could occur are unlikely to do so. This is an enor-
mous task. So much information is available that each search would take
a significant length of time. The task is made easier, but less reliable,
because much information is not available in a computer readable form.
Despite this, the time for each search is significant, and there are a great
many to do. It is also possible that a key reaction has been performed in the
past, but was not included in the available on-line databases, or else an
erroneous result is recorded as if it were true.
Second, it is hard to define similarity in this context. A synthesis may
require the transformation of an alcohol to a ketone, and there is ample
literature precedent for this. But if there are other alcohols in the mole-
cule, or other groups of atoms which may be affected by the same condi-
tions, it may not be possible to establish this from the literature. If an
alcohol is in an unusually crowded position, it may be rather hard to
change it into a ketone. Literature precedent may include some crowded
alcohols, but nothing quite as crowded, or nothing quite as crowded in the
same way. This may be because nobody has tried a similar reaction, or it
may be that similar reactions have been tried but found not to work. In
the latter case, the unsuccessful result may not have been recorded in the
literature.
For both these reasons, a strategy based simply on literature searching
is unlikely to be competitive with the best synthetic chemists, who would,
of course, use the literature to aid their synthetic designs. It may seem,
then, that organic synthesis will remain a skill in which computers cannot
compete with humans for some considerable time to come. However, this
is not necessarily so.
Information technology enables computers to know the chemistry lit-
erature better than any person, but this, in itself, is not sufficient to design
syntheses of new compounds. The use of information technology, coupled
with methods for the computational analysis of novel reactions, may
enable computers to design better syntheses.
The development of the WWW has shown that it is possible for com-
puters to communicate on a global scale, and this, coupled with develop-
ments in theoretical chemistry, may lead to computers making useful
contributions to synthetic strategies in the near future. The internet has