279 F. Supp. 2d 1238 | S.D. Ala. | 2003
UNITED STATES of America,
v.
Kevin Layne BROWN and Ronald Keith Brown, Defendants.
United States District Court, S.D. Alabama. Southern Division.
*1239 Carlos A. Williams, Christopher Knight, Federal Defender's Office, Gordon G. Armstrong, III, Mobile, AL, for Defendants.
Donna Barrow Dobbins, George F. May, U.S. Attorney's Office, Mobile, AL, for Plaintiff.
Kevin Layne Brown, pro se.
Ronald Keith Brown, pro se.
MEMORANDUM OPINION and JUDGMENT
BUTLER, District Judge.
This matter comes before the court on the bench trial of Defendants Kevin and Ronald Brown, with all the factual elements of the criminal act with which they were charged having been stipulated in a Joint Stipulation of Facts (Doc. 145), with the exception of the issue of whether the chemical structure of 1,4 butanediol ("BD") is substantially similar to the chemical structure of gamma-hydroxybutyric acid ("GHB") within the meaning of 21 U.S.C. § 802(32)(A) and § 813.
I. BACKGROUND
Kevin and Ronald Brown were both charged in a superseding information with conspiracy to distribute and possess with intent to distribute a controlled substance analogue, BD, knowing that the substance was intended for human consumption, in violation of 21 U.S.C. §§ 813, 841(a)(1) and 846. "A controlled substance analogue shall, to the extent intended for human consumption, be treated, for the purposes of any Federal law as a controlled substance in schedule I." 21 U.S.C. § 813. The Analogue Act defines a "controlled substance analogue" as a substance:
(i) the chemical structure of which is substantially similar to the chemical structure of a controlled substance in schedule I or II;
*1240 (ii) which has a stimulant, depressant, or hallucinogenic effect on the central nervous system that is substantially similar to or greater than the stimulant, depressant, or hallucinogenic effect on the central nervous system of a controlled substance in schedule I or II; or (iii) with respect to a particular person, which such person represents or intends to have a stimulant, depressant, or hallucinogenic effect on the central nervous system that is substantially similar to or greater than the stimulant, depressant, or hallucinogenic effect on the central nervous system of a controlled substance in schedule I or II.
Id. at § 802(32)(A).
II. SUBSTANTIAL SIMILARITY
The court finds it appropriate to first address whether Section 802(32)(A) should be read disjunctively or conjunctively since the Eleventh Circuit has never ruled on the issue. See United States v. Fisher, 289 F.3d 1329, 1338 (11th Cir.2002). Obviously, if the statute is read in the disjunctive, then the Defendants are guilty because they have stipulated that BD, once ingested and metabolized, has an effect on the central nervous system that is substantially similar to or greater than the effect on the central nervous system of the controlled substance, GHB. See Joint Stipulation at ¶ 3. The court notes that the majority of courts, in dealing with this issue, have read the statute in the conjunctive which requires the government to prove clause (i) and either clause (ii) or (iii). See, e.g., United States v. Hodge, 321 F.3d 429, 432-39 (3d Cir.2003) (surveying cases and going into extensive detail on the subject). Since a conjunctive or disjunctive interpretation of the Analogue Act is not the primary issue here, the court, for the purposes of the order, assumes that the statute should be read in the conjunctive.
Turning to the issue at hand, the Defendants contend that the term "similar in chemical structure" is not defined by statute and should be explained as it is used in the scientific community. Doc. 136. "[W]here Congress has used technical words or terms of art, `it (is) proper to explain them by reference to the art or science to which they (are) appropriate.'" Corning Glass Works v. Brennan, 417 U.S. 188, 201, 94 S. Ct. 2223, 41 L. Ed. 2d 1 (1974) (citations omitted). Section 802(32)(A)(i) does not use the term "similar in chemical structure" as the Defendants contend; rather, the phrase reads "substantially similar to the chemical structure." There is no dispute that the term "chemical structure" should be explained as it is used in the field of chemistry, but the Defendants and the government are at odds over whether the phrase "substantially similar" should be interpreted according to its common or scientific meaning.
Words that have both a technical and common definition are construed in the latter sense unless the statute indicates otherwise. Huffman v. C.I.R., 978 F.2d 1139, 1145 (9th Cir.1992). Since the Analogue Act does not indicate that the term "substantially similar" is to be defined as it is used scientifically, the court will interpret those words as they are used in everyday language. The dictionary defines "similar" as "1: having characteristics in common: strictly comparable 2: alike in substance and essentials ... 3: not differing in shape but only in size or position." MERRIAM WEBSTER'S COLLEGIATE DICTIONARY 1093 (10th ed.1994). "Substantial" is defined in relevant part as "1a: consisting of or relating to substance b: not imaginary or illusory ... c: important, essential ... 2b: considerable in quantity: significantly great ... 5: being largely but not wholly that which is specified." Id. at *1241 1174. Therefore, these definitions of the words "substantial" and "similar" will be the definitions employed by the court in determining whether the chemical structure of BD is substantially similar to the chemical structure of GHB.
Furthermore, the court is guided by controlling precedent in defining the language "substantially similar" found in the Analogue Act by its common meaning. In United States v. Carlson, 87 F.3d 440, 443-44 (11th Cir.1996), the Eleventh Circuit, in finding that the phrase "substantially similar" in the Analogue Act was adequately defined, relied on the holding of United States v. Hofstatter, 8 F.3d 316, 322 (6th Cir.1993). Carlson, 87 F.3d at 444. In Hofstatter, the Sixth Circuit held that "a penal statute must define the offense with sufficient definiteness to enable ordinary people to understand what conduct is prohibited and must do so in a manner that does not encourage discriminatory enforcement." Hofstatter, 8 F.3d at 321 (citations omitted). The Sixth Circuit then found that the language "substantially similar" was "sufficiently precise to enable an ordinary person in the position of [the defendants] to know that listed precursor chemicals ... should not be possessed for the purpose of manufacturing, for human consumption, substances similar to [controlled substance analogues]." Id.
Likewise, in United States v. McKinney, 79 F.3d 105 (8th Cir.1996) overruled on other grounds by 520 U.S. 1226, 117 S. Ct. 1816, 137 L. Ed. 2d 1025 (1997), the defendant argued that the Analogue Act was unconstitutionally vague because there was no scientific consensus that the chemicals at issue were controlled substance analogues. McKinney, 79 F.3d at 108. The defendant asserted that "if experts disagree as to whether the chemical structure of one drug is substantially similar to a controlled substance, then the statute is unconstitutionally vague as to that drug." Id. The Eighth Circuit rejected the defendant's argument and instead focused on how a reasonable layperson would have examined the two chemical structures. Id.
In our case, a reasonable layperson could, for example, have examined a chemical chart and intelligently decided for himself or herself, by comparing their chemical diagrams, whether the chemical structure of two substances were substantially similar. At trial, two experts testified that aminorex and phenethylamine were analogues under the statute, and one expert apparently drew diagrams of phenethylamine and methamphetamine for the jury's comparison. We have examined the charts that appellant has submitted and believe that they would have put a reasonable person on notice that the substances in question were substantially similar within the meaning of the statute.
Id. Therefore, it is apparent to the court that the phrase "substantially similar," as contained in the Analogue Act, does not hinge on how a scientist would interpret the term, but how ordinary people use the language.
Three courts have dealt with the two substances at issue here, BD and GHB. See United States v. Washam, 312 F.3d 926 (8th Cir.2002); United States v. Niemoeller, 2003 WL 1563863 (S.D.Ind. Jan.24, 2003); and United States v. Roberts, 2002 WL 31014834 (S.D.N.Y.2002). In Washam, the defendant argued that the Analogue Act was vague as it applied to him because there was no scientific consensus that BD's chemical structure was substantially similar to GHB's chemical structure. Washam, 312 F.3d at 930. The Eighth Circuit again defined the term "substantially similar" as a person of ordinary intelligence would. Id. at 930-31. The Circuit Court found the phrase sufficiently definite:
*1242 The term "substantially similar," as used in the statute, does not mean "exactly the same." There obviously will be differences in chemical structures between an "analogue" chemical and a schedule I or II chemical. If two chemicals' structures are exactly the same, the chemical in question would no longer be an "analogue," but instead it would be the same chemical as the listed chemical. Thus, some level of difference is acceptable between an analogue's chemical structure and a proscribed chemical's structure.
Id. In Niemoeller, the defendant alleged that the Analogue Act was unconstitutionally vague because there was no clear scientific meaning of the term "substantially similar." Niemoeller, 2003 WL 1563863, at *2. The district court found that the phrase did not render the statute defective because it put persons of average intelligence on notice of the proscribed conduct even though it did not provide absolute certainty to a person seeking to "experiment in reaching the outermost boundaries of lawful conduct[.]" Id. at *4.
However, in Roberts, the experts on both sides argued about the importance of functional groups in determining structural similarities, and the importance of displaying the molecules as a two- or three-dimensional model. Roberts, 2002 WL 31014834, at *2. In holding that the Analogue Act was vague, the district court found that ordinary people have no chance to discern the meaning of the phrase "substantially similar to the chemical structure" if scientists could not even agree on the proper methodology that should be used to determine structural similarity. Id. at *4. The district court reasoned that even if diagrams of the molecules "were made available to a layperson, the lack of consensus by experts in the field as to the import of those diagrams demonstrates that they could not provide such a person with the degree of notice sufficient to know whether their conduct would be prohibited by the Analogue Statute." Id.
Here, in interpreting the statute, the court finds that Congress in passing the Analogue Act never intended for the public to have to apply a chemical formula to determine if a substance was an analogue. Had they, Congress could have provided in Section 802(32)(A)(i) languagesuch as "the chemical structure of which has substantially similar molecular structures or chemical properties to the chemical structure of the controlled substance ..."that would have required scientific evidence to not only prove what the chemical structures were of the two substances, but whether the molecular structures and/or chemical properties were substantially similar as well. Rather, Congress simply said the chemical structures (as explained by an expert) had to be substantially similar, thus making the comparison between the two far easier to quantify in less exacting terms than a chemical analysis between the two. The fact that the government's expert witnesses, Drs. DeFrancesco and Irwin, were unable to scientifically "prove" their comparisons between BD and GHB is not a failing in their testimony, but simply a result of the constitutionally sufficient (see Carlson, supra) comparison Congress required.
Furthermore, differing methodology in reaching a conclusion on substantial similarity does not render the statute vague, but may tip the scales one way or the other on the issue in the mind of the trier of fact depending on the relative strength of the analysis. See Washam, 312 F.3d at 931 (stating that all experts need not agree on substantial similarity in chemical structures in affirming a conviction).
Moreover, when deciding the issue of substantial similarity in chemical structure, the metabolization of a compound is a relevant factor that should be taken into *1243 account. The district court in Roberts viewed this as a conflation of clauses (i) and (ii) when it stated that the chemical structure of a molecule must be analyzed without considering its reaction in the human body upon ingestion. Roberts, 2002 WL 31014834, at *5. In disagreeing with the holding in Roberts, this court is mindful of controlling precedent and the purpose of the Analogue Act. First, in Fisher, the Eleventh Circuit stated, "It is also undisputed that upon ingestion, GBL [a controlled substance analogue] converts into a GBL metabolite: GHB. Therefore, this Court finds that GBL upon ingestion meets the definition of a controlled substance analogue as its chemical structure and effect on the central nervous system are substantially similar to GHB, a Schedule I Controlled Substance." Fisher, 289 F.3d at 1339. Implicitly, it appeared that the Eleventh Circuit deemed the starting point for comparing the two structures to be upon ingestion and not on a blackboard. Second, in order to allow a sensible approach to the Analogue Act's concerns about human consumption of these compounds, other courts have stated that metabolization and effects on the human body under the second part of the statute may be considered when determining substantial similarity of chemical structure under the first part of the statute. See Washam, 312 F.3d at 932-33; Niemoeller, 2003 WL 1563863, at *5 ("consideration of structural similarity does not or should not take place in a vacuum").
Finally, the Defendants further assert that the rule of lenity requires that the criminal statute be narrowly construed here as the term "substantially similar" is ambiguous. Doc. 136. The court has already determined that "substantially similar" is not ambiguous and is to be given its everyday meaning. "The rule of lenity ... is not a doctrine of first resort whenever a criminal defendant identifies a potential ambiguity in a statute ... Instead, the rule of lenity applies only when `the traditional canons of statutory construction' fail to resolve an ambiguity." United States v. Maldonado-Ramirez, 216 F.3d 940, 943 (11th Cir.2000) (citations omitted). While the term "substantially similar" is flexible enough to change with the circumstances, it is not ambiguous and for that reason the Defendant's rule of lenity argument must fail.
III. FEDERAL RULE OF EVIDENCE 702
In Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993), the Supreme Court reaffirmed that the district court judge is to screen scientific evidence for relevance and reliability; otherwise, the jury might be exposed to confusing and unreliable expert testimony. Daubert, 509 U.S. at 597, 113 S. Ct. 2786. "While these concerns are of lesser import in a bench trial, where no screening [for] the factfinder can take place, the Daubert standards of relevance and reliability for scientific evidence must nevertheless be met." Seaboard Lumber Co. v. United States, 308 F.3d 1283, 1302 (Fed.Cir.2002). Federal district courts are still "required to rely only on admissible and reliable expert testimony, even while conducting a bench trial." Gonzales v. National Bd. of Med. Exam'rs, 225 F.3d 620, 635 (6th Cir.2000) (Gilman, J., dissenting) (listing cases). However, "district courts conducting bench trials have substantial flexibility in admitting proffered expert testimony at the front end, and then deciding for themselves during the course of trial whether the evidence meets the requirements" of Rule 702. Gonzales, 225 F.3d at 635. Alternatively, in a bench trial, it has been an acceptable method "to admit evidence of borderline admissibility and give it the (slight) weight to which it is entitled." See SmithKline Beecham Corp. v. Apotex *1244 Corp., 247 F. Supp. 2d 1011, 1042 (N.D.Ill. 2003).
Rule 702 of the Federal Rules of Evidence provides:
If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case.
FED. R. EVID. 702. Daubert set forth a non-exclusive checklist for trial courts to use in assessing the reliability of scientific expert testimony. Daubert, 509 U.S. at 593-95, 113 S. Ct. 2786. One of the factors noted by the Daubert Court is whether the technique or theory has been generally accepted in the scientific community. Id. at 594, 113 S. Ct. 2786. Although the Supreme Court in Daubert also listed other factors, no attempt has been made to codify them, and they are neither exclusive nor dispositive. FED. R. EVID. 702 (advisory committee notes).
First, for good reasons, both government witnesses, Drs. James DeFrancesco and Richard Irwin, were stipulated to by the Defendants as experts in their respective fields of chemistry. Dr. DeFrancesco has been employed as a forensic drug chemist with the Drug Enforcement Agency ("DEA") for the past six years. Gov't ex. 1. He holds a B.S. in Chemistry from Elmhurst College, a Ph.D. in Physical-Organic Chemistry from Michigan State University, and has accumulated an impressive amount of work experience in the private sector and academia. Id. Dr. Richard Irwin has been employed as a biochemist/toxicologist with the National Institute of Environmental Health Sciences ("NIEHS") for the past twenty-three years. Gov't ex. 10. He holds a B.S. in Chemistry from Baldwin Wallace College, a Ph.D. in Biochemistry from the University of Tennessee, performed Post-Doctoral work at the Johns Hopkins University, and also has accumulated an impressive amount of work experience in academia. Id. Furthermore, he has authored a peer-reviewed paper entitled "National Toxicology Program Summary Report on the Metabolism, Disposition, and Toxicity of 1, 4-Butanediol." Gov't ex. 9.
Both doctors, in determining the chemical structures of BD and GHB and assessing the similarity of the two chemical structures, viewed both molecules in their two-dimensional static state and formed opinions based on their knowledge, skill, and experience. Both doctors stated that a visual assessment was the best method of forming an opinion on structural similarity and is generally accepted in the scientific community. Furthermore, both doctors stated that the rapid metabolization rate of BD into GHB factored into their opinion. Without more, the court finds that both doctors had sufficient facts and data before them in forming their opinions. Likewise, both doctors employed an accepted method in reaching their conclusion and applied it reliably to the facts before them. Accordingly, the court holds that the opinions of both doctors on the similarity of chemical structures between BD and GHB are both relevant and reliable.
The Defendants' witness, Dr. John Steele, has garnered a B.S. and M.S. in Botany from Auburn University, a Ph.D. in Plant Pathology from the University of Minnesota, and performed Post-Doctoral training at the University of Wisconsin. Def. ex. 1. From his testimony at trial, Dr. Steele's academic work was primarily in *1245 the field of plant pathology and botany, with his Ph.D. thesis dealing with the estrogenetic changes in swine reproductive systems caused by a certain fungi in corn. During his course of study, he claimed to have placed an emphasis on chemistry in general. Id. He began his professional career as a research associate in the field of plant pathology for nine years at the Universities of Minnesota and Wisconsin, and moved into the private sector mainly as a scientist for twenty-two years. Id. Dr. Steele, at trial, stated that he has worked with GHB and BD on isolated projects even though he does not possess a license to work with controlled substances. Dr. Steele now earns his living as a freelance consultant.
The court is unwilling to qualify Dr. Steele as an expert with knowledge in the area of controlled substances, especially BD and GHB. However, because of the relaxed nature of the Daubert standards in a bench trial, the court will still consider Dr. Steele's testimony, but his lack of expertise in the area substantially affects the weight of his testimony. However, even if the court were to qualify Dr. Steele as an expert in this matter, for the reasons set forth below, the court finds that he did not use reliable methods in reaching his conclusion, and the methods he did employ were not reliably applied.
IV. BURDEN OF PROOF
As an element of the offense, the government has the burden of proving beyond a reasonable doubt that the chemical structure of BD is substantially similar to the chemical structure of GHB. See United States v. Klecker, 228 F. Supp. 2d 720, 727 (E.D.Va.2002).
A. Determination of Substantially Similar
First, the court found the two government witnesses to be well-qualified on the subject matter. All three witnesses generally defined "chemical structure" as an arrangement of atoms within a molecule and the way they are bonded to one another. As chemists, both government experts stated that they used the term "substantially similar" just like a layperson. Dr. DeFrancesco went on to define "substantially similar" in several ways: not identical; the majority of the chemical features that are shared are similar; or the two chemical structures are almost the same but not quite. Dr. Irwin stated that the definition of "substantially similar" would mean very similar.
However, Dr. Irwin admitted that the term "substantially similar" in the scientific arena would depend on the context of the discussion; that there was no hard-line definition of the term; that there was no way in which to quantitatively assess similarity without resort to intuition because such a comparison defies quantification; and that the definition of "similar chemical structure" depends on the judgment of a chemist. He testified that if six chemists were asked about their opinion of the substantial similarity between two chemical structures, then six different answers may be offered. "`Substantial similarity' is a gut-level decision.... And all `gut feelings' are legitimate." Likewise, Dr. DeFrancesco admitted that he is unable to quantify the percentage of similarity between BD and GHB. He conceded that other molecules may be more similar to GHB than BD, but that there were tens of millions of organic compounds and he did not have a cut-off point in dividing similar chemical structures from substantially similar chemical structures. However, while pinpoint scientific quantification on the issue of structural similarity is helpful, it is the trier of fact that makes the ultimate decision from the evidence submitted. Moreover, not every issue in a case in which scientific experts testify requires a scientific formula.
*1246 The relevant formulas are as follows:[1]
Gov't ex. 6A. Both chemicals are simple linear molecules with a four carbon chain. The main difference between GHB and BD is at the terminal carbon. The differing carbon in BD is referred to as a tetrahydron carbon, while in GHB it is referred to as a trigonal planer. Also, while both molecules are flexible, GHB is more likely to fold over on itself whereas BD is more likely to stay in linear form. Likewise, the extra oxygen atom in GHB converted the alcohol functional group, as found in BD, into a carboxylic acid functional group.[2] Finally, the extra oxygen atom in GHB is connected to the hydrogen atom by a double bond which changes its inter-nuclear position because of the length of the bond. However, all these differences in structure are attributable to the replacement of the two hydrogen atoms in BD with the one oxygen atom in GHB. Although chemical properties do not relate to structure, GHB is more acidic than BD, but the two molecules are not complex structurally, so one change in structure may have disproportionate effects on the physical and chemical properties of the molecule.
Once ingested, the chemical structure of BD converts into the chemical structure of GHB in a two step process within minutes.[3] This metabolic process is the natural reaction in the body to BD and cannot be prevented. First, after BD is ingested, it reacts with alcohol dehydrogenase ("ADH"), an enzyme, in converting to gamma hyroxybutyraldehyde ("GHBH"). Second, GHBH reacts with aldehyde dehydrogenase ("ALDH"), another enzyme, in converting to GHB. Again, the net effect is that the terminal carbon atom in BD sheds its single bonds to the two hydrogen atoms and forms a double bond with an oxygen atom. In the process, the length of the bonds connecting to the terminal carbon change during the conversion. The conversion can be summed up as follows:
*1247
Gov't ex. 7.
Despite the dissimilar functional groups on one end of the shared carbon chain, both government experts opined that BD was substantially similar in chemical structure to GHB. Moreover, both government witnesses explained the significance of the metabolization process on their analysis of chemical structure. Dr. DeFrancesco noted that the conversion process contributed to his conclusion as a bio-chemist since the modification of a single carbon was a fairly fast process in this instance. He further stated that the body's rapid conversion of the structure of BD into GHB indicates that the rest of the molecule is the same. Dr. Irwin testified that the rapid conversion process factors into his opinion that the chemical structures of BD and GHB are substantially similar since the alcohol functional group and carbon chain of both molecules have to be similar in order to bind in the body. He noted that the common functional group and carbon chain in both molecules serve as a "handle" in BD, allowing enzymes to attach and convert the molecule into GHB. Although he explained that basic metabolization of one molecule into another does not generally mean that the two are structurally similar, in this case he opined that the rapid conversion emphasized the similarity in structure. He stated that the two molecules would have to be structurally similar because of the reaction of the body's receptors to the identical "handle" found in both BD and GHB.
The entire transformation of BD into GHB takes place on the terminal carbon and only involves the replacement of two hydrogen atoms with one oxygen atom. Although other changes, such as bonding and differing functional groups, are attendant to this conversion, they are based solely on the change at the terminal carbon. Notwithstanding the defense characterization of the government experts' opinions as "gut feelings," the court finds a user's involuntary and rapid "gut reaction" of converting BD into GHB as more persuasive evidence that BD and GHB are substantially similar in chemical structure. The chemical "handle" of BD remains unchanged after the molecule is converted into GHB, further emphasizing the substantial similarity in chemical structure between the two molecules. In reaching the conclusion that the chemical structures of BD and GHB are substantially similar, the court, as will be further explained below, completely discounts the testimony of Dr. Steele.
B. The Tanimoto Coefficient and Computer Database Searches
Notwithstanding the above determination that the chemical structure of BD is substantially similar to the chemical structure *1248 of GHB beyond a reasonable doubt, the court finds it necessary address the testimony of Dr. Steele, the defense witness. Dr. Steele criticized the government's experts claiming that they only used impressionistic, non-quantitative methods which usually are not considered scientific. However, when asked if it is an acceptable method to simply look at the linear structures of BD and GHB in assessing similarity, he stated that he was not aware of the methods of other chemists. He then attempted to quantify the structural similarity between BD and GHB in two ways.
First, Dr. Steele employed the Tanimoto coefficient[4] to determine the percentage of similarity in chemical structure between BD and GHB. He noted that this method was more reliable and effective than just looking at structures because it had the advantage of not being subjectively weighted. He stated that under the Tanimoto coefficient BD was 46.7% similar to GHB. Accordingly, Dr. Steele opined that since the percentage is below 50%, then BD must not be substantially similar to GHB because "substantial" means largely or a good deal above 50%.
In a move that would make Las Vegas odds-makers and Enron accountants proud, Dr. Steele unnecessarily compared the entire structures of BD and GHB with each other in calculating the Tanimoto coefficient of the substructures of the two molecules. The net effect of this manipulation was to skew the equation before the comparison of the substructures even started.[5] Dr. Steele attempted to justify his method by stating that substructures alone do not suffice to distinguish one chemical from another. However, the court found this answer unsatisfactory since two distinct molecules may be entirely comprised of the same substructures which are nonetheless arranged differently. Further, Dr. Steele claimed that he needed the intact molecule in his analysis because he did not want to lose information in the comparison; although, he conceded that he did not have the proof he needed to back up his claim. Regardless of Dr. Steele's suspect techniques, the court determined that the Tanimoto coefficient, as used in this case, impermissibly emphasized the differences between BD and GHB by counting the COOH compound found in GHB, but not BD, several times. The skewed results convinced the court that the Tanimoto coefficient is not a reliable indicator of structural similarity when assessing two simple linear molecules with only one differing functional group, nor was it even reliably applied in this instance.
Second, Dr. Steele ran several computer database searches using three different software programs, Chemfinder, ChemIDplus, *1249 and Advanced Chemical Development ("ACD"), to compare the structural similarity of BD and GHB.[6] The first software program, Chemfinder, uses the Tanimoto coefficient to calculate structural similarity between compounds. Obviously, the court, for the above reasons, finds the results of Chemfinder to be unpersuasive as the software is based on the Tanimoto coefficient. Under all the searches using the two other programs, ChemIDplus and ACD, neither BD nor GHB showed up as being 50% similar to the other compound. Thus, Dr. Steele, on the basis of these results, concluded that BD and GHB were not substantially similar to one another in chemical structure.
However, some of the results from the searches were bizarre and curious. For instance, under ChemIDplus, the compound 4-Phenylbutylamine came up with a similarity rating of 51.40% to BD. Gov. ex. 11. On rebuttal, Dr. Irwin stated that he would be stunned if someone claimed that 4-Phenylbutylamine was more similar than GHB in chemical structure to BD. GHB and BD both have oxygen atoms while 4-Phenylbutylamine has no oxygen atoms. Gov. ex. 12. 4-Phenylbutylamine has a NH2 group and a benzene ring, consisting of six carbon atoms, four hydrogen atoms, and three double bonds between some of the carbon atoms within the ring, while BD and GHB do not have a NH2 group or benzene ring. Id. As illustrated, the two-dimensional chemical formula of 4-Phenylbutylamine reads:
Id.
Also, in another query under ChemIDplus, the compound butanoic acid came up with a similarity rating of 52.31% to GHB. Gov. ex. 11. Again, on rebuttal, Dr. Irwin stated that he would be extremely surprised if someone claimed that butanoic acid was more similar than BD in chemical structure to GHB. Butanoic acid consists of a benzene ring and a chlorine atom, while BD and GHB do not contain these substructures. Gov. ex. 13. As illustrated, the two-dimensional chemical formula of butanoic acid reads:
Id. Finally, the court notes that GHBH, the intermediate molecule in BD's conversion to GHB, did not rate as being 50% similar to GHB despite the apparent likeness between the two molecules. Gov. ex. 11.
In assessing the reliability of these results, the court tried to ascertain how ChemIDplus and ACD were programmed. The government experts and Dr. Steele all agreed that the term "similarity" had to be measured in some way to allow the software to search for matches. Therefore, the programmer necessarily had to leave his or her fingerprints on the algorithm, or the set of instructions, which allows the programs to perform certain functions such as assessing the structural similarity of the molecules. Dr. Steele conceded that he did not know the parameters of the relevant software beyond the guess that it probably used some variation of the Tanimoto *1250 coefficient along with other factors. He admitted that he did not know what the other factors might be because the parameters of the software were "proprietary" information. Both government experts noted that, according to the results, the programs appeared to be taking chemical properties or functional groups into account. Furthermore, both government experts testified that they did not believe that other scientists would use computer database searches for comparing the structural similarities of two known simple molecules. Dr. Steele conceded that he did not know if these types of searches were commonly used for this purpose. In Dr. Steele's own words, "My measurements are simply others I ran in programs, the parameters of which I do not know.... I simply ran the numbers."
Obviously, without knowing how the computer programs define "similarity," the results are of little, if any, value to the issue at hand. The court finds it amazing that Dr. Steele would stake his professional reputation on the results of a program without knowing how it works. Furthermore, in solely relying on the search results as being the undisputed measure of structural similarity, the stubborn refusal by Dr. Steele to supplement his opinion by visually assessing 4-Phenylbutylamine, butanoic acid, BD, and GHB, reeked of bias. Complete acceptance of another person's subjective definition of "similarity" does not make for an objective scientific method; rather, it is blind reliance devoid of merit. Accordingly, the court is unable to find that the database searches, as used here, were reliable, much less reliably applied.
V. CONCLUSION
Based on the expert opinions of both government witnesses along with the stipulated facts, the court finds beyond a reasonable doubt that the Defendants, Kevin Layne Brown and Ronald Keith Brown, are GUILTY of conspiracy to distribute and possess with intent to distribute a controlled substance analogue, BD, knowing that the substance was intended for human consumption, as charged in Count One of the superseding information.
Sentencing for the Defendants will be held on October 16, 2003, at 9:30 a.m. in Courtroom 5A.
NOTES
[1] The letters represent atoms (C=carbon, O=oxygen, and H=hydrogen). The dash between the sub-structures represents a bond while the number two represents the number of hydrogens in the substructures.
[2] Functional groups refer to the way in which atoms are bonded in giving the molecule a chemical property.
[3] The court found the ambiguous language contained in the DuPont sales literature for terathane to be unconvincing in trying to impeach the government's evidence on this point. See Def. ex. 11.
[4] The Tanimoto coefficient works by parsing out all the different combinations of substructures contained in the two molecules being compared. With both molecules, one would start on one end and cut off the first substructure from the molecule, then cut off the first two connected substructures and proceed accordingly stopping after the next to last substructure on the opposite end. Likewise, one would then repeat the process starting on the opposite end. Finally, one would start in the middle of a molecule and work out toward the ends stopping after the next to last substructures on both ends. After counting up the total number of substructures, one would divide the number of shared substructures by the total number to reach the percentage of similarity. See Def. ex. 4.
[5] Here, Dr. Steele found that BD and GHB shared 7 common substructures out of a total number of 15, making BD 46.7% similar to GHB. (7 ÷ 15 = 46.7%). Def. ex. 4. However, when the intact molecules of BD and GHB are removed from the equation, leaving just the substructures, both molecules share 7 common substructures out of a total number of 13, which makes BD 53.8% similar to GHB. (7 ÷ 13 = 53.8%).
[6] The concept behind these programs is straightforward and simple. The target compound being searched is entered into the query field, and the software searches the database for compounds it deems similar, by a desired degree or percentage, to the target compound.
