Search This Blog

Wednesday, November 26, 2014

Ketchum Sample 26, The Smeja Kill: Independent Lab Reports

The sample of flesh and hair collected by Justin Smeja weeks after his reported sasquatch killings has been analyzed for DNA sequences by four labs:
1.  As part of the Ketchum DNA study (Sample 26) by her associates.  They sequenced the entire mtDNA genome and 2.7 M  nDNA bp.
2.  By Wildlife Forensic DNA Laboratory at Trent University, Ottawa, submitted by Tyler Huggins (see link at right).  They sequenced the human HV1 mtDNA region and black bear STR nuclear microsatellite loci.
3.  By DNA Solutions, submitted by Bart Cutino.  They sequenced the human HV1, HV2 and black bear cytochrome b mtDNA regions.
4.  By Mitotyping Technologies, submitted by Prof. Bryan Sykes as part of his recent paper (see link at right).  They sequenced the black bear 12S rRNA mtDNA gene.
In addition, the Ketchum results were reinterpreted by me, but I had no sample and did no new laboratory experiments (See my three papers at right) .

All three independent labs (2., 3., 4.) concluded that the sample was from a black bear, as did I based on the published nDNA  sequence in the Ketchum paper.  Labs 2. And 3. also found human contamination.  Prof. Sykes cleaned up his sample first, and so saw no human mtDNA, only black bear.  This would seem to put the nail in the coffin, except that Melba has said (e.g. to me privately) that Justin Smeja sent different known bear samples out in place of Sample 26, so that the results would exonerate him from possible criminal charges for killing something resembling a human.  Is there any way we can check on this claim?  Not exactly, but we can show that the human mtDNA in all three samples is virtually the same – Justin Smeja’s.
Related to this, Scott Carpenter in his blog, “Ketchum DNA Study - Sample 26” ( made the claim that Melba found H1a as the haplogroup, while Huggins (Scott thought) gave haplogroup A, and Cutino T2 for the human contamination.  Can a sample have three different human haplogroups?  Carpenter said no, and therefore the original sample, Ketchum 26, could not have been split to Huggins and Cutino.  But are these different haplogroups accurately determined?   Unfortunately not.  Each has a problem. 



Scott is correct in that the Huggins report (see at right) lists haplogroup A for the controls and the sample in its Table 1.  However, further into the report (p. 2) details conflict with haplogroup A, which is First American out of Asia: “Further analysis indicates that it is a European haplotype; it occurs with highest frequency in East Europe (11%) and Caucasus (10%) (i.e. it initially originated from near the  Caucasus  mountains  regions  between  the  Black  and  Caspian  Seas).  It is  not  known  to  have originally occurred in East Asia, Southeast Asia, Australia, Oceania (i.e. New Zealand, Papua New Guinea, etc.), North America, South America or Central America.”  I asked Tyler Huggins about this discrepancy, and he said that the Table 1. listing “A” was a place holder meant to indicate that all four samples, Huggins_1 (the flesh), 570, 576, and 578 (the last three buccal swabs from Justin Smeja) had identical haplogroups, not Haplogroup A.  This is poor reporting on the part of the lab and confusing.
The Huggins report indicates that the HV1 region of mtDNA (positions 15999-16400) matched 100% NCBI database accession JQ705199.  Now we have the sequence which we can compare to Cutino’s published HV1 and to the Ketchum S26 mtDNA sequence.  The results are as follows:
Compared with rCRS (revised Cambridge Reference Sequence), the extra HV1 mutations are:
Huggins  16126C  16187T   16294T  16296T  16304C             
Cutino    16126C   16187T   16294T  16296T  16304C
Ketchum  16126C  16187T  16294T  16296T  16304C

Thus, Huggins and Cutino samples are identical, and Ketchum differs from these two by the one highlighted mutation 16313T.

Huggins did not sequence HV2, but we can compare Cutino and Ketchum:
Ketchum has one extra HV2 mutation 384G.
How significant are these two extra mutations (16313T and 384G)?  16313T occurs nowhere in the most recent version (Build 16) of the mtDNA Phylotree (link at right).  384G occurs only twice (haplogroups T1a2a and U5b1h) in the entire Phylotree, which is based on over 20,000 different human mtDNA sequences.   My complete mtDNA sequence interpretation of the Ketchum samples (my second paper at right) showed that Sample 26 was outside the range of known human haplotypes by 16 extra mutations including 384G and 16313T.  The average number of extra mutations is less than 3 (2.37 based on my subsample of 35 H1a database entries).  Based on statistics of mutation frequencies, the Ketchum sample had a 1 in 224,056,304 chance of being from the known human population; therefore, it has no human haplogroup.  Justin Smeja’s three buccal swabs matched the Huggins sample over HV1 and therefore also the Cutino sample   






Let’s go back to the reported Haplogroups:
Huggins: A (known to be false – see above)
Cutino: T2
Ketchum: reported H1a but equally likely to be H5e – both with 16 extra mutations, not the “one novel SNP” claimed in her paper, Table 2.
The Cutino determination T2 was based on the HV1 mutations 16126C, 16294T, and 16304C only; add 16187T and 16296 and the haplogroup is T2b3e, which however does not match the coding region mutations of the Ketchum sample (coding region was not determined by Huggins or Cutino). Ketchum said the sample had “one novel SNP” from H1a.  It has 16.  A sample this far from human cannot be correctly haplogrouped, as I pointed out in my second paper.     
One more piece of data, the Cutino report included a cytochrome b sequence, which matches black bear best and other bears next, no human.  How can one sample produce human mtDNA (HV1, HV2) and bear mtDNA (cyt b)?  It’s the primers – human specific for HV1, HV2 and universal for cyt b, respectively.  The sample is black bear with human impurity.
So, with high probability all three samples had Justin Smeja’s mtDNA over the HV1 region.  All three as well as one more (Sykes’) tested positive for black bear.  Scott, you have to dig into the details more by doing BLAST™ searches and comparisons.  Your analysis was too shallow to reveal the truth.  It relied on contract lab interpretations and misstatements. And unfortunately, contract labs need very close monitoring to get good results.  As one Ketchum Peer Reviewer said, “It is very difficult to manage the quality control and release standards of contract laboratories from the outside.”  Here, one (Huggins’) made misleading/incorrect table entries (haplogroup A), another (Cutino’s) ignored extra mutations (16187T, 16296T) without comment or qualifiers, and a third (Ketchum’s) reported a Haplogroup H1a without commenting on the 16 extra mutations which make it improbably human.

Thanks to Scott Carpenter for raising the issue and making the reports available on his website (link at right).  But sorry, Scott, your conclusion is wrong.   NO THANKS to Melba Ketchum who refuses to release the Smeja Sample 26 for further testing (e.g. as outlined above by me), after repeated public and private requests by Derek Randles, who submitted it to her.  What is she hiding?  Respectable scientists cooperate with efforts to cross check their work; the best even solicit it and collaborate. Regardless of their legitimate points of disagreement, when the results come out differently, they don’t impugn the integrity of other investigators without proof of any wrong doing.
My take on this sample is that it has no verifiable connection to the Smeja killings. Only Justin Smeja and his hunting companion know what he shot, and it has no fresh, connected, and documented sample to analyze.  Even matching Smeja’s DNA in the various samples doesn’t prove they were all from a single original, only that he contaminated them all. What this issue really needs is for Melba Ketchum to repeat the Huggins STR analysis with black bear primers or release the sample for others to do so.  Then we can see whether the black bear in both samples is the same bear.   I asked her to do this early last year and she ignored me, or at least has not published the results.

Now for the shocker.  The STR analysis with black bear primers in the Huggins report Figures A7.1 and A7.2 shows more than two peaks (alleles; normally there’s only one from each parent) for each gene, so the bear portion is also contaminated, probably by a second bear, but possibly by
Justin’s dog (more research needed here).   Did two bears fight over the sasquatch carcass, one biting or clawing off a piece of flesh from the other and contaminating it?  In the process did the Sample 26 also get contaminated with original sasquatch DNA, confusing the whole mtDNA sequencing by Ketchum, which turned out to be improbably human and possibly a mixed result?