Jump to content

The Ketchum Report (Continued)


Guest Admin

Recommended Posts

Guest VioletX

And another from Dr K's F/B:

Melba Ketchum

There are so many arm chair scientists out there criticizing everything so I thought I would address one of the more serious accusations....that of other species contaminating our samples. With a novel species, random sequence will sometimes be obtained with primers meant to amplify a certain locus in a known species (since that locus may not have the same flanking sequences in the novel species). This can either yield "junk DNA" or a coding sequence elsewhere in the unknown species' genome. When you then BLAST this novel sequence, it can BLAST as similar to some known species, especially if what you have ended up amplifying is not a coding region. This is also true when it comes to human whole genomes. If you just take a random sequence from the raw sequences obtained by next generation sequencing, that very well may be "junk DNA" rather than a coding region and you can get all kinds of other species coming up in the BLAST. Folks should learn their genomics prior to criticizing.

Link to comment
Share on other sites

My understanding of Phred scores is that they measure the possibility of artifacts or errors in sequencing, not the possibility of contamination. In other words, it looks at the likelihood that your dataset has things like missing bases or incorrect base pairs, e.g. AC or TG pairings. At least, that's what I remember from scientific evidence.

Could someone with more knowledge of the area please clarify?

Exactly right a Q score calls out how accurate and sound a base call is. it has nothing to do with the reassembly of the genome, nor does it rule out contamination. A contaminated sample could easily have a very high Q scoere, it's only after the many very short (300 BP) are assembled , thats when we start looking at what those sequences tell us. And in this case, according to what very little data Dr Ketchum published, her data is clear - it is bear contaminated with human. seems like we have heard that somewhere before.

As for the segments that do not align, go look up the effects of bleach and formaldehyde have on DNA strands, then with that information, understand how the genome sequence is reconstructed from the very short pieces, then recognize that they started at human chromosome 11 and built out!

Link to comment
Share on other sites

This is so funny! They keep on going on and on about the Q30 score, and even lifted some info from the illumina web site regarding Phred scores which take into account errors in the sequencing (resulting form incomplete chemistry during the cycling resulting in the sequence of some of the reads being out of register one behind or one early, as leisureclass mentioned). I have spent the last couple of days reviewing the technical literature on the HiSeq2000 regarding this and even went to ask the technician who runs our machine here. There is no indication that I have found that the Q30 score would be influenced by a mixed population in the sequencing library. The Q30 score is based on each cluster, which is derived from one single DNA sequence!!!! It seems like there will be a few rebuttal posts to this ridiculous statement!

Edited by ridgerunner
Link to comment
Share on other sites

And another from Dr K's F/B:

Melba Ketchum

There are so many arm chair scientists out there criticizing everything so I thought I would address one of the more serious accusations....that of other species contaminating our samples. With a novel species, random sequence will sometimes be obtained with primers meant to amplify a certain locus in a known species (since that locus may not have the same flanking sequences in the novel species). This can either yield "junk DNA" or a coding sequence elsewhere in the unknown species' genome. When you then BLAST this novel sequence, it can BLAST as similar to some known species, especially if what you have ended up amplifying is not a coding region. This is also true when it comes to human whole genomes. If you just take a random sequence from the raw sequences obtained by next generation sequencing, that very well may be "junk DNA" rather than a coding region and you can get all kinds of other species coming up in the BLAST. Folks should learn their genomics prior to criticizing.

I'm sure that Leonid Kruglyak appreciates Dr. Ketchum's advice.

Leonid not Leonard. Sorry Doc.

Edited by leisureclass
Link to comment
Share on other sites

BFF Patron

Contanamation, contanamation, is making me wait......... :music:

Link to comment
Share on other sites

And another from Dr K's F/B:

Melba Ketchum

There are so many arm chair scientists out there criticizing everything so I thought I would address one of the more serious accusations....that of other species contaminating our samples. With a novel species, random sequence will sometimes be obtained with primers meant to amplify a certain locus in a known species (since that locus may not have the same flanking sequences in the novel species). This can either yield "junk DNA" or a coding sequence elsewhere in the unknown species' genome. When you then BLAST this novel sequence, it can BLAST as similar to some known species, especially if what you have ended up amplifying is not a coding region. This is also true when it comes to human whole genomes. If you just take a random sequence from the raw sequences obtained by next generation sequencing, that very well may be "junk DNA" rather than a coding region and you can get all kinds of other species coming up in the BLAST. Folks should learn their genomics prior to criticizing.

As one of those so called armchair scientists, I am going to make this statement personal. MK needs to take her own advice. wow... I will need to critique this at a later point...

Link to comment
Share on other sites

Guest craichead

Just wanted to chime in on the Q30 scores aspect and how it supposedly shows it's pure DNA:

I've spent about an hour on the internet researching Q30 scores and I can't find a single source that claims a Q30 score can determine the purity of a DNA sample. Not one. I found one comment by a genetecist that said that Q30 scores tell you nothing about purity, only errors in sequencing.

I believe this is the case and that the Q30 argument is being used out of context. I don't think it shows anything about whether or not the samples were cross contaminated. If someone can show me otherwise please do.

Link to comment
Share on other sites

Melissa: He put this up as "required viewing"

I haven't been able to watch it all yet, but the first 7 min or so is nothing but snickering and jeering and the FB comments are mostly more of the same.

I saw that someone copied one of my posts from here and posted it in the comments for this FB video. That wasn't me that did that.

Link to comment
Share on other sites

BFF Patron

I think it is Tontar :keeporder:

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...