150 MUTATIONS 



In each of the induced spectra there is, of course, some spontaneous 

 background which has to be subtracted out. 



Magni: In this map have you not yet subtracted the spontaneous 

 mutants? 



Benzer: I can't, because I don't know which ones they are, but 

 in analyzing it statistically, you have to not take seriously recur- 

 rences at a site where they can be largely due to spontaneous mutation. 

 On the other hand, if you have a site where there are 8 or 10 recur- 

 rences with nitrous acid and only one spontaneous out of 20 times as 

 many mutants, the probability is extremely high that that is really a 

 nitrous acid hot spot. 



Lederberg: I think it would help if you indicated what weight to 

 give the different lines, because they do represent different mutants. 



Benzer: Yes, the number of mutants on each line is different. It is 

 very difficult for us here to assess the significance at the various 

 points. It would take too long. If you look at the coincidence of hot 

 spots, i.e., look at a point which is a hot spot for nitrous acid, and ask 

 whether or not it is a hot spot for other mutagens, you can give each 

 such spot a set of quantum numbers, if you will. Unfortunately, a 

 clear decision may be limited by the statistics of how many mutants 

 you have actually studied. 



You find, for instance, with the ethylmethane sulfonate spectrum, 

 there are 26 of what statistically ought to be significant hot spots, 

 and, out of these, 19 are also present in the nitrous acid spectrum, 6 

 are also present in the aminopurine spectrum, 5 are present in the 

 diaminopurine spectrum, 7 are present in the bromouracil spectrum, 8 

 are present in the deoxybromocytidine spectrum, none in proflavine. 



One can pick out many, perhaps 25, different kinds of spots having 

 different responses, different sets of decisions as to whether or not 

 they are hot for such or such a mutagen. 



Goodgal: Do you have a sufficient number of hot spots with each 

 of the mutagens so you can make — 



Benzer: These are the ones I could say something about. The num- 

 ber might be much larger if one studied more mutants. 



Atwood: Your matrix may become simpler if you have more. 



Benzer: A spot being not hot is also statistically significant. 



Atwood: I see. You have omitted those. 



Benzer: I call a spot cold if you expect three or five mutations there 

 and find zero. Of course this can be due to chance. Coldness is also 

 subject to statistical test. 



Lederberg: Twenty-five? Does that mean there were eight different 



