1. In general, I was unable to read the DQAI and PM typing, strips provided to me, as the photocopies were of poor quality. Nevertheless, I attempted to evaluate the interpretation of the FSA's results using, the information I had available to me (e.g., some of the photocopies of typing strips were of good enough quality to read, and the majority of typing results were documented in hand written notes).

2. The only apparent oversight regarding the reporting of results is for item A-59-1. Given the mixture observed, the FSA cannot discount the possibility of a 1.2 DQAI allele.

3. Assuming that the typing, strips accurately reflect the reported results, the conclusions drawn by the FSA, specifically in their two consultation reports, are accurate at face value; i.e., Richard Eberling- could not be excluded as one of the donors of the DNA typing, results obtained from specific items of evidence (i.e., stain from Marilyn Sheppard's bed sheet, and items A-59-1, A59-2, b-3-b-1, and 1-C). However, a "failure to exclude" could mean anything- without some means to convey the weight of the typing results. The FSA does not provide a statistical interpretation of their results in order to assess the weight, or meaning of their conclusions. This must be done by the FSA at trial if their results and conclusions are to be entered into evidence. Otherwise the conclusions may be prejudicial to the triers of fact.

4. The DQAL and PM profiles reported by the FSA for specific items of evidence (i.e., stain from Marilyn Sheppard's bed sheet, and items A-59-1, A59-2, b-3-b-1, 3, and 1-C) are mixtures of more than one individual. In fact, some of the items are a mixture of three or more individuals. For example, the DQAI/PM profile of item A-59-1 (sperm fraction of a vaginal smear from Marilyn Sheppard), has six (6) possible DQAI alleles (i.e., each numerical designation indicates the presence of another "form" or allele of DQA1, where each allele is a different DNA sequence). An individual can only have two (2) DQAI alleles, thus, there are at least three individuals represented in this sample. If one were to assess the weight, or meaning of the fact that Richard Eberling- could not be excluded as a donor of DNA to this mixture, one could assess how many individuals in the general population could have contributed an allele to the mixture. However, the statistical methods to make this assessment can be somewhat complex. On the other hand, one could deter-mine how many individuals in the population could not have contributed DNA to the mixture by simply determining the percentage of the population that has alleles not represented in the mixture. The following is an example of this approach for item A-59-1:

The information provided in the above table illustrates that the DQAI profile reported for item A-59-1 could be shared by 91.7% of the African American population, or 97.2% of the Caucasian population. Even more compelling is that only 8.3% of the African American or 2.8% of the Caucasian population could be excluded as a possible donor of the mixture in item A-59-1, and that 100% of both populations can be expected to share the same PM profile. Thus, while Richard Eberling, cannot be excluded as a donor, which may in fact be consistent with other circumstantial evidence, the DNA results provide little additional information to support the hypothesis that he was a contributor.

8. It would be useful to know what validation studies were conducted to allow the FSA to perform re-amplification of certain DQAI and PM reactions. Proper studies are required to ensure the reliability of the results. Of greatest concern is that the positive control contains both a 1.1 and 4.1 DQAL allele, which could be a source of contamination during reamplification. In defense of the FSA, it does appear that negative controls were clean throughout the testing process.