Well, just a note of some interesting articles regarding new statistical techniques to analyze sequence libraries in a manner akin to analyzing DGGE profiles:
Lozupone, C. and R. Knight. 2005. UniFrac: a New Phylogenetic Method for Comparing Microbial Communities. AEM: 71: 8228-35.
And an application of UniFrac:
Ley, R.E. et al. 2005. Obesity alters gut microbial ecology. PNAS. 102:11070-11075.
To use UniFrac, you will need to download and operate "Python" software.
You will also need to download a special "Numeric" module. Make sure you download the Numeric, not the NumPy, module.
Then you will need to download the .zip file of the unifrac program (http://bayes.colorado.edu/unifrac.zip).
The UniFrac program is not trivial to use if you haven't had any experience with Python or similar languages. I've emailed Dr. Lozupone and she has indicated that eventually a web-based UniFrac program will be available. However, not yet.
You will need phylogenetic trees in Newick tree format as the input data file for this program.
Tuesday, January 17, 2006
Monday, November 21, 2005
DGGE Help Page
I am announcing the release of my DGGE Help Guide. I have spent some time reviewing the questions that have accumulated on the YahooGroups DGGE page and decided to compile a guide to answer many of the questions that come up repeatedly. The guide is not meant to be definitive, but it is meant to be a resource for users to find the information that they do need. I also put in some of my own personal opinions regarding PCR and DGGE.
I hope that people find the guide useful. I will update the guide as time goes along, and I will be happy to receive additional information from anyone interested in providing it. If anyone feels that there are sections that they would like to write, update or correct, please contact me.
I have the guide on another blog: http://ddgehelp.blogspot.com/
The entire text can be had in word format or PDF format. However, I have found that the PDF format does not retain the hyperlinks. For the intended format, please use the word file.
I hope that people find the guide useful. I will update the guide as time goes along, and I will be happy to receive additional information from anyone interested in providing it. If anyone feels that there are sections that they would like to write, update or correct, please contact me.
I have the guide on another blog: http://ddgehelp.blogspot.com/
The entire text can be had in word format or PDF format. However, I have found that the PDF format does not retain the hyperlinks. For the intended format, please use the word file.
Friday, November 11, 2005
Monthly Update
Well, it seems about time for a monthly update. I had thought I might do this more routinely, but I guess this is what we can expect! I have added a few tidbits to the SuPER PCR page. These include citations to two very nice manuscripts. One is a method similar to SuPER PCR but for RNA and the other is a clever mechanism to circumvent the problem of plastid DNA in environmental studies of plant roots.
Things here are going well. I am in the midst of a major sequencing effort relating to the "Sulfate II" experiment that was conducted at NASA. The idea was to take hypersaline microbial mats (See Bebout et al. 2005; Minz et al. 1999 and many others) and incubate them under controlled conditions at standard salinity (85 ppt), lowered salinity (35 ppt), at standard sulfate concentrations (~60 mM) and at low sulfate concentrations (< 1 mM). After letting these mats acclimate to these conditions, a suite of geochemical measurements were taken (oxygen/sulfide profiles, methane evolution, etc), and I have been working, together with Jason Smith (U of Florida), Mary Hogan (NASA/ARC), Brad Bebout (NASA/ARC) and Lee Bebout (NASA/ARC), on analyzing the methanogens and sulfate reducers in the system. The preliminary analyses of the methanogens are quite striking and indicate a dramatic shift in population as a result of the removal of sulfate. More on that when all the data are in. No sequence data yet on the sulfate reducers.
I have also been performing some phylogenetic analyses with the program Mr.Bayes (discussed in a prior posting). However, I am having problems with the technique - with regard to the branch lengths that the method is yielding. I have noticed that the branchlengths produced by MrBayes are significantly longer than those I get from other treeing methods such as Neighbor Joining or Maximum Likelihood. I have observed that the branch lengths are roughly 2x those of NJ, and 2x thes equence difference (as estimated by BLAST or other aligning method). I realize that MrBayes branch lengths are "expected changes per site" and not necessarily the same as NJ branch lengths. Still, the 2x factor seems excessive. If anybody has any thoughts on this I would love to hear them. Overall, the branching order yielded by Mr. Bayes seems quite consistent with other methods.
Well, more soon.
Things here are going well. I am in the midst of a major sequencing effort relating to the "Sulfate II" experiment that was conducted at NASA. The idea was to take hypersaline microbial mats (See Bebout et al. 2005; Minz et al. 1999 and many others) and incubate them under controlled conditions at standard salinity (85 ppt), lowered salinity (35 ppt), at standard sulfate concentrations (~60 mM) and at low sulfate concentrations (< 1 mM). After letting these mats acclimate to these conditions, a suite of geochemical measurements were taken (oxygen/sulfide profiles, methane evolution, etc), and I have been working, together with Jason Smith (U of Florida), Mary Hogan (NASA/ARC), Brad Bebout (NASA/ARC) and Lee Bebout (NASA/ARC), on analyzing the methanogens and sulfate reducers in the system. The preliminary analyses of the methanogens are quite striking and indicate a dramatic shift in population as a result of the removal of sulfate. More on that when all the data are in. No sequence data yet on the sulfate reducers.
I have also been performing some phylogenetic analyses with the program Mr.Bayes (discussed in a prior posting). However, I am having problems with the technique - with regard to the branch lengths that the method is yielding. I have noticed that the branchlengths produced by MrBayes are significantly longer than those I get from other treeing methods such as Neighbor Joining or Maximum Likelihood. I have observed that the branch lengths are roughly 2x those of NJ, and 2x thes equence difference (as estimated by BLAST or other aligning method). I realize that MrBayes branch lengths are "expected changes per site" and not necessarily the same as NJ branch lengths. Still, the 2x factor seems excessive. If anybody has any thoughts on this I would love to hear them. Overall, the branching order yielded by Mr. Bayes seems quite consistent with other methods.
Well, more soon.
Thursday, October 6, 2005
Contract Renewal
Well, good news here. My NRC contract was renewed for a second year. It's nice to be employed. However, since the NRC fellowship is essentially limited to two years, it is time to start looking for another position. I have applied for some grants as well, but it will be quite a while before I hear about the results of those applications.
Things are starting to get quite busy. I have a number of sequencing projects to support a large amount of geochemical data acquired in a long-term project here at Ames. The projects involved adding nutrients (ammonia,nitrate or phosphorus) to microbial mats maintained in the greenhouses here, and a second major experiment in which salinity and sulfate levels were modulated. So, we are examining major microbial functional groups: methanogens, sulfate-reducers, photosynthetic populations, methane oxidizers, and nitrogen-fixers. More updates on the results of these experiments soon.
Also, happy new years (shana tova) and happy birthday (to me).
Things are starting to get quite busy. I have a number of sequencing projects to support a large amount of geochemical data acquired in a long-term project here at Ames. The projects involved adding nutrients (ammonia,nitrate or phosphorus) to microbial mats maintained in the greenhouses here, and a second major experiment in which salinity and sulfate levels were modulated. So, we are examining major microbial functional groups: methanogens, sulfate-reducers, photosynthetic populations, methane oxidizers, and nitrogen-fixers. More updates on the results of these experiments soon.
Also, happy new years (shana tova) and happy birthday (to me).
Friday, September 16, 2005
GenomiPHI Update
I've done a little more work with the GenomiPHI kit. It is not a panacea for all that ails with PCR done directly on environmental DNA. I have had mixed results with different samples. Some samples with weak PCR do not fare well with the kit while others do. I have also determined that halving the recommended volumes does not help - you should use their recommendations.
On the plus side, I performed a denaturing gradient gel electrophoresis (DGGE) analysis of an environmental DNA extract directly and after pre-amplification with the genomi-PHI kit. The bacterial 16S ribosomal RNA gene profiles look identical before and after pre-amplification, suggesting that the pre-amplification is not biased.
Wednesday, September 7, 2005
GenomiPHI
Well, it's been a little while since I've updated. These things happen. I've been playing a bit with this full genome amplification kit called GenomiPHI (www.genomiphi.com). The kit contains a DNA polymerase (Phi29) and random primers. The polymerase is quite nice because it is less sensitive to environmental contaminants (humics, etc) than can be co-extracted with DNA. It works at low temperatures (30 deg C), does not require thermal cycling and is a proofreading enzyme.
The enzyme is useful because it makes copies (up to 5000x) of total genomic DNA. Apparently this is an unbiased reaction (I haven't tested this myself). I have so far employed this technique to enhance PCR amplification of the mcrA gene from methanogenic archae (see below). The kit is about $150 for 25 reactions - not cheap, but not out of reach.
In high sulfate environments such as marine microbial mats, methanogens do not compete well with
sulfate-reducing bacteria and are generally not abundant. Thus, the detection of methanogens by PCR with mcrA gene primers can be difficult. I tried to circumvent this problem by performing a pre-amplification with the genomiPHI kit. I took several DNA extracts from high sulfate mats and diluted them either 1/10th or 1/100th. 1 ul of these dilutions were then added to the genomiPHI reaction with random primers and run overnight at 30C. It turned out that diluting the initial DNA sample was not a great idea. The first figure shows the DNA after genomiPHI amplification on a 0.8% agarose gel (first 7 lanes, left to right) and a standard DNA extract (lane 8).
No obvious difference between dilutions was observed after the overnight pre-amplification (the first three lanes are the same sample, undiluted, 1/10 and 1/100th). These samples were then subject
to standard PCR with mcrA primers. The first seven samples (left to right) are the PCR yield from the reaction containing DNA pre-amplified with the genomiPHI kit. The next 3 are the same samples amplified without pre-amplification. The next 1 (strong yield) is a PCR yield from a low-sulfate mat sample generating high levels of methane. The final lane is the negative control that unfortunately has a slight bit of contamination. Anyway, sample #1 yielded strong PCR amplification when the undiluted DNA extract was subject to GenomiPHI pre-amplification. At a 1/10th dilution the PCR yield was weaker and at a 1/100th dilution, no PCR product was detected even after pre-amplification. The same sample gave only a weak PCR yield when amplified directly from the original DNA extract. A dilution of the DNA extract prior to the pre-amplification turned out to be a poor idea, and these samples did not amplify well (except for #16) even after GenomiPHI amplification. More work to be done, but still, it is interesting.
If interested, check out this article: Gonzalez et al. 2005. Multiple displacement amplification as a pre-polymerase chain reaction (pre-PCR) to process difficult to amplify samples and low copy number sequences from natural environments. Environmental Microbiology. 7(7):1024-1028.
The enzyme is useful because it makes copies (up to 5000x) of total genomic DNA. Apparently this is an unbiased reaction (I haven't tested this myself). I have so far employed this technique to enhance PCR amplification of the mcrA gene from methanogenic archae (see below). The kit is about $150 for 25 reactions - not cheap, but not out of reach.
In high sulfate environments such as marine microbial mats, methanogens do not compete well with
sulfate-reducing bacteria and are generally not abundant. Thus, the detection of methanogens by PCR with mcrA gene primers can be difficult. I tried to circumvent this problem by performing a pre-amplification with the genomiPHI kit. I took several DNA extracts from high sulfate mats and diluted them either 1/10th or 1/100th. 1 ul of these dilutions were then added to the genomiPHI reaction with random primers and run overnight at 30C. It turned out that diluting the initial DNA sample was not a great idea. The first figure shows the DNA after genomiPHI amplification on a 0.8% agarose gel (first 7 lanes, left to right) and a standard DNA extract (lane 8).No obvious difference between dilutions was observed after the overnight pre-amplification (the first three lanes are the same sample, undiluted, 1/10 and 1/100th). These samples were then subject
to standard PCR with mcrA primers. The first seven samples (left to right) are the PCR yield from the reaction containing DNA pre-amplified with the genomiPHI kit. The next 3 are the same samples amplified without pre-amplification. The next 1 (strong yield) is a PCR yield from a low-sulfate mat sample generating high levels of methane. The final lane is the negative control that unfortunately has a slight bit of contamination. Anyway, sample #1 yielded strong PCR amplification when the undiluted DNA extract was subject to GenomiPHI pre-amplification. At a 1/10th dilution the PCR yield was weaker and at a 1/100th dilution, no PCR product was detected even after pre-amplification. The same sample gave only a weak PCR yield when amplified directly from the original DNA extract. A dilution of the DNA extract prior to the pre-amplification turned out to be a poor idea, and these samples did not amplify well (except for #16) even after GenomiPHI amplification. More work to be done, but still, it is interesting.If interested, check out this article: Gonzalez et al. 2005. Multiple displacement amplification as a pre-polymerase chain reaction (pre-PCR) to process difficult to amplify samples and low copy number sequences from natural environments. Environmental Microbiology. 7(7):1024-1028.
Thursday, August 11, 2005
16S rRNA Gene Sequence Aligner (Online)
Here's an online sequence aligner for 16S rRNA gene sequences that I heard about from Ruth Ley at the MBL workshop. It solves one of the problems with the ARB aligner - namely that the ends of the sequences are scattered all over the place and you have to manually adjust them.
The only downside is that this aligner aligns the sequences to other databases such as Phil Hugenholz's. Still, nice to avoid the problem with the ends.
http://greengenes2.lbl.gov/cgi-bin/nph-NAST_align.cgi
The only downside is that this aligner aligns the sequences to other databases such as Phil Hugenholz's. Still, nice to avoid the problem with the ends.
http://greengenes2.lbl.gov/cgi-bin/nph-NAST_align.cgi
Subscribe to:
Posts (Atom)
