There are 2 errors in what I posted below about the forensic analysis discussed in Professor Jacques Ravel's slides. The first correction is that the over 1,100 samples submitted to the FBI were screened for genotypic differences, and may not have been screened for morphologic differences.
The second error is that the slides say wild-type Ames (isolated from a heifer that died in Texas in 1981) did not vary from an anthrax sample isolated from Bob Stevens' cerebrospinal fluid (over 5 million base pairs identical), rather than from the anthrax letters directly.
What I termed slide 16 has a list of 5 different isolates (morphotypes) from the Leahy letter and 3 from the NY Post letter. One of each is termed "wild-type" so you might say there are 4 (Leahy: A, B, C and E/Opaque) and 2 (NY Post: A and B) morphotypes distinct from the majority colony type listed on the slide.
Later the slides indicate that 3 genotypes were identified within the A morphotype. The slides do not indicate how many genotypes were found in the other morphotypes. Nor how these were determined. There are 2 slides on each page. So what I identified as slide 16 is actually page 16, slide 31.
Slide 45 indicates that 4 PCR assays were developed to distinguish genotypes. I don't know from which samples these distinctive genotypes were found, and used.
The over 1100 submitted samples were screened for PCR genotype differences. Some had 1 or 2 variants but only the sample from Ivins' flask had 4. (It is implied that 4 were found in the anthrax letters.)
I erred in mixing up morphotypes and PCR assays in my earlier statement. There is nothing in the slides that indicated the 1100 samples were screened for morphologic colony differences.
There is missing information so I cannot say what kinds of samples were used to discern the genotypic differences. But my guess is these were genetic differences within morphotypically distinct colonies.
However, the slides appear to indicate that the morphologic variants differed between the two letters.
If genotypic differences of rare morphotypes are what was used to nail Ivins' flask, the science may be questionable. If the morphotypes varied between the two letters, is it forensically valid to focus on qualitative genotypic differences in a subset of morphotypes? But I am guessing this is what was done. The FBI needs to tell us precisely what methods were used.