Molecular Ecology - Science topic

Molecular Ecology, 3rd Edition

Joanna R. Freeland

ISBN: 978-1-119-42615-8 January 2020 384 Pages

For the mantel test, you need to use the arp file attached here

Thanks

Hi, Lara!

Maybe I can help.

If I understood well, you are using a 96 well plate and you place this well into a PCR machine? And after that, you seal it with a silicone mat? And your PCR protocol has a hot-start period where you need to activate the polymerase enzyme?

If everything mentioned above is the case, then I am afraid that using these silicone mats could, unfortunately, be the problem. Especially during the initial activation, where you are holding the samples at a high temperature for a longer time and the pressure is building up inside each well, resulting in the silicone mat popping out.

Additionally, this can be an issue due to the 96 PCR well being too smooth as well. But I wouldn't worry about the PCR plate as it is important that it is smooth so that DNA doesn't stay on it.

There are several ways how to avoid this:

1) Try to buy some other sealing mats to see if they work better with your 96-well plate. Or try buying new mats and plates together from the same company if you already didn't do that.

2) Use a PCR kit and protocol without a hot start, although I wouldn't recommend that as it could potentially affect your primer specificity and lead to primer dimers.

3) Try to buy standard 0,2 mL PCR strips for 8 samples in a row. Something like this:

I use such strips for my PCR as I have more control with opening and closing these tubes and using them at once only as much as I need. These strips never disappointed me and never popped out while performing PCR, and I did it with them in 5 different PCR machines that our lab has to offer.

I hope this was of some help.

Best regards,

Domagoj

Hi Larissa

I'm having the same problem as you had, with 5 loci.

I tried to change the Data type: Standard, and use the haplotype data, not the sequence, but I'm not sure if it works the same with the sequence.

Yes, ape in R is a good option

For the sake of providing a direct answer to the SPECIFIC question you asked, the following R code will do exactly what you require, and output the tree to a PDFfile in your current working directory

library(ape)

tree = read.nexus(file="path\to\file")

tree_export = "tree.pdf"

pdf(file=tree_export, width=6, height=6)

plot(tree, cex =0.5, use.edge.length=FALSE)

axisPhylo()

edgelabels(cex = 0.25, width = 0.1)

dev.off()

depending on your tree structure just play around with the "cex"and "width" parameters to make the branch numbers readable, the above peramters are reasonable starting points.

The above code assumes your tree is in Nexus format. For the commonly used Newick format do this:

library(ape)

tree = read.tree(file="path\to\file")

tree_export = "tree.pdf"

pdf(file=tree_export, width=6, height=6)

plot(tree, cex =0.5, use.edge.length=FALSE)

axisPhylo()

edgelabels(cex = 0.25, width = 0.1)

dev.off()

i.e. swap "read.nexus" for "read.tree" function

Hi Benoît,

No program that I am aware of. I reckon that differentiate life-strategist from mere amplicon seqs can be, at best, unreliable if considering problems such as uncomplete reference datasets.

A good paper describing this but trying to list taxa that behaves as copio-oligotrophs:

Please let me know if you come across with a relevant program though

Cheers

Ben

Yes this question can really only be addressed in a helpful manner with more information about your needs and unfortunately your budget. But basically eDNA techniques are emerging that allow simultaneous amplification/sequencing of all of the OTUs that successfully amplify, so by definition this approach uses universal primers for markers usually such as 16S rRNA and ITS, internally transcribed spacer region, or COI cytochrome c oxidase I, then the resulting sequences are blasted against NCBI database to determine biodiversity in the sample. But if you are using a more traditional PCR approach where you want to specifically amplify only certain target lineages, one possibility is use species of specific primers. Depending on the species of interest, this can be tricky if such PCR primers are not available, but you can design your own primers that specifically amplify only the species of interest, provided there are DNA sequences available for these taxa.

How much of the sequence is published? If its just the microsatellite region, then you might be out of luck. If there is sequence flanking, you can design your own.

This website might give you some initial lights:

Also, note it has a arXiv paper associated to it.

I had some troubles running diveRsity::writeBoot.

diveRsity::basicStats and diveRsity::diffCalc ran my data. However, I had to make minor adjustments to my data to writeBoot (remove some empty spaces).

Also, parallel argument must be set to TRUE. parallel = FALSE causes problems.

Hi Ana,

In the scenario you specify "t2 varNe 2 N2" but "N2" is already the effective population size specified on line 1. So this line is not really changing the Ne of population 2 like you might want it to. Try changing it's name to see if this solves your problem. Also make sure the naming of all variables is consistent. You have t1-db in the scenario but just t1 in the conditions.

Also, for your conditions you can also remove some redundancies. For example, t3>t2 and t4>t3 imply that t4>t2 and so you could remove this condition. For the same reasons N2>N1b and t4>t1 are redundant conditions and could be removed as well.

Hope this helps!

A paper worth reading (linked below) illustrates that the time lag between an isolation barrier and detecting this pattern via genetic clustering methods (eg. STRUCTURE or TESS) can be as low as 15 generations for organisms with relatively large dispersal abilities. In other words, impacts to genetic clusters can be altered relatively quickly, which is especially pertinent to your study organism. Hope this helps.

For those with access to a Mac with OSX, the following may be helpful, as it appears the Mac version is updated more often. I was able to get NewHybrids (without graphics) to run reliably using the steps below. Note that I'm no programming expert, so there may be easier ways, but this worked well for me after trudging through available documentation online. Hopefully this saves someone some hours of agony.

1) Download and extract all files from GitHub site: github.com/eriqande/newhybrids

2) Follow folders system to Bin -> OSX -> newhybridsng.exec (run this; no graphics version)

3) Back at the main folder downloaded from GitHub, open a new Terminal Window at this folder location (can be done on some machines by right-clicking the main folder)

4) Copy and past the following into the terminal to see a helpfile of all available commands: ./bin/OSX/newhybsng --help-full

5) Using this documentation, string together necessary analysis of desired datafile. Note that "--no-gui" command is needed at end to make the "no graphics" version work. The below example loads the test data file "TestDat.txt" that comes with NewHybrids, along with personalized model specifications. Also note that your data files are likely stored in a different location, but you may be able to drag-and-drop the data file into your Terminal window to obtain the file path needed.

Example Terminal command: ./bin/OSX/newhybsng --datafile /Users/Andrew/Desktop/newhybrids-master/test_data/TestDat.txt --theta-prior uniform --pi-prior uniform --burn-in 50000 --num-sweeps 100000 --no-gui

A combination of haplotype grouping, network analysis of genes interaction and population genetic background of affected paitients could produce interesting results. We have initiated this type of study with male infertlity in our country, I hope in comming 2 years we can get some introductory results and insights to this hypothesis.

This paper is very helpful

Depends on the number of loci and quality of the data. IMA2 and DiYABC requires many loci and high quality of the data. Use different bottleneck tests. To check for contemporary bottlneck you can try Wang's Colony. 

Hi, 

For dominant markers like ISSR, you can determine many genetic diversity parameters like Nei gene diversity, Shanon information index, effective No. of alleles, etc. Please read the paper attached and other papers published by our team. 

You van use Genalex and Popgun for such analyses. 

Regards 

Masoud Sheidai 

using 100 SNPs (assuming a reasonable minor allele frequency) is not such a difference in resolution compared to classical AFLP studies. 

gene diversity is nothing else than the expected heterozygosity (He). If you have an average gene diversity of c. 0.15 - 0.20 (which is the chance that if you take two random alleles from your sample they are different), you're in the typical range for this type of population genetic studies. 

In the early days, many allozyme studies had similar diversity ranges (but fewer loci). So it's pretty OK to use He, compare it to Ho, calculate allelic richness, and so on. The more loci, the smaller the error on the estimate of genome-wide diversity and so on. So more loci allow you to discern small effects from noise better. 

Basically, all classical population genetics is based on detecting deviations from Hardy-Weinberg expectations (F-statistics at different hierarchical levels) to test for evolution. (the HW theorem assumes no evolution, hence it is used as a null model). Any deviation from HW is caused by evolution, and the trick is to disentangle the three (four) major drivers of evolution: drift, gene flow (and subdivision), selection (which is different per locus due to recombination) and mutation (which is in itself typically negligible compared to the three other).

More importantly, what questions are you trying to answer? Do you require individual-based analyses (assignments, clustering, parentage, ...), are you satisfied with estimates at the population level (summary statistics), etc.

As a reviewer, I'd rather read a manuscript with a good sampling design for the question asked, crisp statistical analyses, clear knowledge on the assumptions of each analysis (the biggest issue in population genetics if you ask me), and using only 100 SNPs, than a study that uses 100 000 SNPs but has no clue how to really answer the question (other than "we were interested in ..."), other than by blindly copying what others have done in somewhat similar studies.  

@Ivan and @Yann: Thanks for comments. I am using Freebayes to call SNPs because it has an argument that considers the mapped BAM files as deriving from pooled samples. BUT I REALLY WANT HAPLOTYPES!

Also, I have worked with the Popoolation tools before, but my understanding is that they are better suited to projects where you have much larger contigs (e.g. choromosomes as opposed to RAD fragments).

It confuses me that so little has been done to achieve this gap in analysis tools.

Barcode Gap is a suitable method to delineate species from a set of aligned sequences as long as you have enough sequences per species. Yet, it is important to remind that Barcode Gap should not be the only criterion! Complementary evidences like other genetic markers, morphology and distribution are necessary to make a decision. I recommend reading the ABGD method, see link attached.

Once you have established a reference dataset with barcode gap then you can be confident with the identification of an unknown specimen through it's barcode sequence.

Hope that helps.

Thanks Bruce for the very helpful answer about Mass Sectrometry.

Kind regards

Timothy

Hey James,

Not sure what you mean by 'corrected allele frequencies' but with BayeScEnv2 you can actually identify the putative loci under selection (I think it calculates local Fst and correlates this measurement of genetic differentiation with the environment). So you can always extract a data set with only the loci under selection and plot them. Is this answering your question?

Yes, you can divide your dataset into partitions and apply different sequence evolution models to those partitions. A common framework might first involve using PartitionFinder (http://www.robertlanfear.com/partitionfinder/) to delineate partitions and their respective "best" models.

Sounds like you are using Mr. Bayes? You would then define your partitions in the Mr. Bayes command block of your nexus file.

For some help in configuring your nexus file for a partitioned dataset, see: http://mrbayes.sourceforge.net/wiki/index.php/Analyzing_a_Partitioned_Data_Set

Biologically, the probability of mutation occurred at the same site more than once (multiple hits) is small (given the low mutation rate), the lower mutation rates the smaller the likelihood given a time interval. Given a certain mutation rate, the smaller the time interval the less likely of substitution saturation.

Hi 

you can use this protocol:

Making a stock solution of 0.4% Trypan Blue (also available for purchase):

Weigh 4 mg of Trypan Blue powder and add 1 mL of water or 1x PBS. Scale up as necessary

 

Using Trypan Blue to stain dead cells in tissue sections:

Mix 1:1 the Trypan Blue solution with 1x PBS

Add 1mL of this dilution to tissue sections, mix well, and incubate at room temperature for 5 minutes

Rinse tissues with 1x PBS for 15 minutes

 

Using Trypan Blue to stain dead cells in suspended cultures:

Dilute your cell suspension in a 1:1 dilution with the 0.4% Trypan Blue stock solution

Carefully fill the hemocytometer chamber with the stained cells

Incubate the cells for 1-2 minutes at room temperature and proceed with counting/quantitation

Hi,

See the link below. 

Good luck. 

Hi Shiva,

Did you find the software? If not how did you calculate PD or PE?

Actually the structuring phase comes after to check the IBD (isolation by distance). For IBD the most commonly use criteria is Mental test but their is also some biases (Please read: Meirmans PG (2012) The trouble with isolation by distance. Mol Ecol 21:2839–284), so I would like to suggest first do IBD through RDA (Meirmans PG (2015) Seven common mistakes in population genetics and how to avoid them. Mol Ecol 24:3223–3231), if their is no IBD then go through the clustering, may be in your case you have  just one genetic group in actual not more than one.

I use reusable plate sealing mat for my PCR. I have not faced contamination problem as far, even after reusing the mat multiple times. 

I would recommend reusable plate sealer but care should be taken while washing and autoclaving to avoid contamination. 

yes there is a "posteriori" assignment test in Geneclass2. I've run it LOOOONG time ago so I'm not sure how it is in the current version, but it wasn't hard to interpret at all. The main question is to have "GOOD reference populations" to assign these potential immigrants to. So what I was suggested to do when I've applied it, is to use as reference pops the genetic clusters resulting from STRUCTURE. Thus, the likely immigrants would be like "ouliers" if you can say that way from those in theory "panmitic" genetic units resulting from Structure. But is fast to run, so you can run both with the genetic clusters and the populations they were captured in....  and interpret the results  from both assignments.

Mr. John,

Thank you for your answer. I will look into heat shock proteins right away. Ideally we might be able to study the entire genome so I will look into that as well!

Many thanks,

Allan Copuyoc

Hey! My question do you want to sequence the SSR product? If yes its fine, I answer your question. Why the question of evolution come here, amazing and interesting answers!! 

What do you mean by linkage inheritance? Genes are linked and are inherited together both in plants and animals.

If you are sure for experimental conditions suggested by Dr. Pandit and Dr.  Ramadevi, It means that your primers targeted more than one loci due to paralog loci . In such a case you can score PCR products desperately for mapping such as SSR1a and SSR1b.  

It depends. For example, if you submit a mitochondrial DNA sequence, call it "Shrew Cytb", but provide no specific information of which position corresponds to the beginning/end of your Cytb, then the sequence will be marked as unverified. What often happens is, you have a partial sequence of Cytb, so you don't fill in the annotation part of it because you think it's not needed. GenBank staff comes back to you asking for this info, but you can't be bothered to do it because your paper needs finalizing, and "why does it matter anyway?" kind of thinking. If that is the case, the sequence will probably be ok, but still marked as unverified.

However an alternative scenario is, someone submits a Cytb sequence will the correct information, but GenBank staff identified it ha a mutations that causes an early stop codon. Because biologically, early stop codons in Cytb will likely be lethal, GenBank staff will ask the author to carefully check the sequence for errors. Again, the author might choose not to chase this for various reasons, and the sequence gets marked as unverified. 

So the short answer to your question is, if you are dealing with unverified sequences from Genbank, always best to make some data integrity checks yourself before using it. It also means you need to be careful if your phylogeny gives some strange and unexpected results. You might find that those strange and exciting results, involve sequences that are unverified. 

Having said that, it's always a good idea to check GenBank sequences for data integrity. There was a very good paper on this a few years back in TREE: http://www.cell.com/trends/ecology-evolution/abstract/S0169-5347(03)00150-2. Obviously, the amount and diversity of data in Genbank has changed considerably since then, but the sentiment remains true. 

Hope that helps.

Hi

You can get binary data, such as for dominant markers like AFLP, into splitstree using this format

#NEXUS

BEGIN Taxa;

 [&&SPECTRONET SN_BLOCKNAME Taxa_2]

 DIMENSIONS NTAX=4;

 TAXLABELS tax1 tax2 tax3 tax4;

END;

BEGIN Characters;

 [&&SPECTRONET SN_BLOCKNAME Characters_1]

 DIMENSIONS NCHAR=10;

 FORMAT DATATYPE=STANDARD GAP=- SYMBOLS = "012345";

 MATRIX

tax1 01010 01010

tax2 11001 00101

tax3 01110 11001

tax4 10101 10101;

END;

For co-dominant markers, you can code the data as binary data by recording the presence/absence of each allele - but this is not a particularly good approach. If you want to combine dominant and co-dominant markers together it will function.

For co-dominant markers alone, it is generally better to use a population genetic software package to generate a distance matrix for the samples/populations and then get that into Splitstree using the format you can find in the examples folder supplied with the software in the file south.nex. Be careful to use an appropriate genetic distance measure in your population genetic software! My experience would suggest being very wary of models that do not treat all alleles as equally related. Typically the Nei (1972) model is a good start.

The following primers give very good results for a large array of bird species:

CO1-Ext Forward: 5’- ACGCTTTAACACTCAGCCATCTTACC-3’

FISH1Reverse: 5’–TAGACTTCTGGGTGGCCAAAGAATCA-3’

annealing temperature=54°C

Refs:

Johnsen, A., Rindal, E., Ericson, P.G.P., Zuccon, D., Kerr, K.C.R., Stoeckle, M. and Lifjeld, D. 2010. DNA barcoding of Scandinavian birds reveals divergent lineages in trans-Atlantic species. J. of Ornithol. 151: 565-578.

Ward, R. D., Hanner, R. and Hebert, P. D. N. 2009. The campaign to DNA barcode all fishes, FISH-BOL. J. Fish Biol.74: 329-356

thanks Melih

Hello Ashley,

It depends on the loci you are using and the group you are examining (hard to generalize across all fish). I have examined several markers in Centrachidae in the southeast US and Catosomidae in the southwest US and here are some insights that I have found.

Both of these groups have very differing substitution rates. Centrachidae has substitution rates similar to that of the average mammal rates, reflecting and more stable history with relatively large populations consent throughout time. Catosomidae, like most western fishes, have had an unstable history driven by volcanism and continental expansion and as such has presumable had periods of bottlenecks and expansions over and over again leading to a quicker substitution rate (almost an order of magnitude), this was found in Chen and Mayden 2012 and is consistent with the work I have in review. So history is very important and other factors including introgression, genome architecture, and genome duplication can also be very important.

Also, as you mention mutation rates vary across the genome. Even 'neutral markers' are not completely neutral but instead nearly-neutral and many of the markers that we call neutral are not. The general rule of thumb for eukaryotes (although these may be more basis to mammals) is an average of 2.2 x 10^-9. Below is a general list for various regions of the genome:

Protein-coding exons - 10^-9

Introns - 10^-8

Ribosomal proteins (like S7 that Derek mentioned) - 10^-6 to 10^-7

Microsatellites (and many other short tandem repeats) - 10^-4 to 10^-6

Mitochondria - 10^-5 to 10^-6

There are also ultra-conservative regions that have a much slow mutation rate and have become popular in the NGS world. Also, when people refer to neutral markers they often are referring to either ribosomal proteins or EPIC markers (exon primed intron coding) that often contain both exons and introns and when you sequence these markers you can often see the rate heterogeneity.

With all that said, applying a standard rate is possible and many researchers do using the standard eukaryote mutation rate of 2.2 x 10^-9 (can not think of the citation right now), but you should be careful and only use it on EPIC markers and not other markers like ribosomal proteins (i.e. S7 and 18S). Also this should be avoided if your organism has had a history of bottlenecks and expansions. The far better thing to do in my opinion is to fossil calibrate your marker as long as you have the calibration points. For examples in fishes see Near et al., 2012 in PNAS for all ray-finned fishes or Unmack et al., 2014 in PloS for Catosomus.

Hope this helps. There is definitely literature out there, they can just be hard to find and typically easier to find if you look more at the family level then in all of fishes.

Best Regards

DARwin 6.0.12

Thanks Tim, 

This paper does look interesting, I will see if I can find more on their website. Thanks also for explaining the turgor loss point, makes more sense to me now.

Cheers,

Stephane.

To begin with, it must be said that you may have some a priori idea about whether a given marker is neutral or has been under selection, however, you cannot know for sure. A coding locus with alleles responsible for different phenotypes could be neutral (phenotypes have or had the same fitness). A microsatellite or other "junk" DNA could be linked to a locus under selection and hitch-hike or have an unknown function and be responsible for changes in fitness. So beware of your assumptions about using a priori neutral and non-neutral markers.

Current widely accepted approach is to consider that the vast majority of the genome behaves neutral and only a handful of loci have been under recent selection. Therefore, we can perform demographic inference from lots of (randomly selected along the genome) markers. If some markers under selection are included, it is assumed that their signal is "crushed" compared to the main neutral signal carried by the overwhelming majority of neutral markers. See however Ewing & Jensen (2016, doi:10.1111/mec.13390).

Once the demographic history of the populations under study has been inferred one can try to detect loci that present diversity pattern incompatible to that history. These can be assumed to have been under selection.

Biochemical methods tests classify them to certain extent, which is just part of identification. You need molecular and advanced molecular methods to identify & classify them. Read below PDF....you would get some idea.

Thanks your answer.

Hi Tyler, 

You might be interested in the papers by Mastretta-Yanes et al (2015) in Molecular Ecology (http://onlinelibrary.wiley.com/doi/10.1111/1755-0998.12291/abstract), and Viricel et al (2014) in Molecular Ecology Resources (http://onlinelibrary.wiley.com/doi/10.1111/1755-0998.12206/abstract)

Cheers, eric

HI,

My templates were bacterial DNA from bird eggshells.

I am working in your suggestions.

Thanks

Urine can be "seen" in the UV helping birds of prey find rodents. Berries differ in their UV reflectance. Some example images of flowers and berries are available at http://www.uv4plants.org/gallery/plants-and-flowers-in-vis-and-uv-images/ as pairs of visible and UV images. Images are from Lasse Ylianttila.

Below are links to some publications I am aware of, but my expertise is more on plants than on animal vision.

SOLVED. There was a non-breaking space in my sample names. Simply removed that and all is well again.

Dear Mr. Phuangphong

In my experience, choosing markers depends on the questions you are planning to answer, and also on the sequence resources already available for comparison. In the case of using different mtDNA regions, as you pointed, usually the choice is based on te available sequences in the databases. COI has been extensively used for insects, and therefore was selected as the region for barcoding especially these organisms. On the other hand, cyt-b was historically used for mammals, and therefore, the database for cyt-B has more data than the COI database for mammals. I am not a specialist in birds, but I would guess that cyt-B and ND3 might have a good database for these organisms.

Hope this helps,

Best

The 2nd option is probably more likely to give a quality answer. If you have the time/computing power I think you should look at the larger dataset, too. If your clustering combines particularly disparate sequences you may be thinning out your taxa distribution and inviting more opportunity for long branch attraction. I would consider comparing the topologies and look for places where lots of relative diversity was collapsed in one area but not in other areas of the tree. If it is even throughout, the smaller dataset is probably much more likely to give a fair answer.

if it's just stored in PBS without fixation, probably not suitable for FISH

The type of barrier varies, but Lilium review by van Tuyl et al. can be a generic pathway, https://www.researchgate.net/profile/Jaap_Tuyl/publication/222482901_Application_of_in_vitro_pollination_ovary_culture_ovule_culture_and_embryo_rescue_for_overcoming_incongruity_barriers_in_interspecific_Lilium_crosses/links/0f3175370713fc92cc000000.pdf

Dear Deborah,

Whether you call these animals Cebuella pygmaeus or Callitrix pygmaea it still remains the same species! Morphology AND molecules seem to indicate that this species is a separated entity within the clade with species of Callitrix. If this is enough to consider it a member of a separate genus is completely dependent of the opinion of the taxonomist: there are no "rules" for this. So you are perfectly entitled to call this species Cebuella pygmaeus (and this might even be the most pragmatic solution). One remark however: do NOT speak about Cebuella when you mean the species; in other words: do not use the name of the genus when you want to indicate a species.

So Johana, you have the choice between the names Cebuella pygmaeus or Callitrix pygmaea for the species that was described for the first time by Spix in 1823 as Iacchys pygmaeus. Because the genus name Iacchus has not been used since, it can be considered a "nomen oblitum" (but the species is valid).

Hi Stevanus,

I found the best method for myself (currently working on extracting DNA from a number of scleractinian families) is to crush the required amount of coral with a mortar and pestle. The skeleton does a good job in lysing the tissue. I usually then store the crushed tissue in a guanidium thiocynate  based "CHAOS" buffer for a couple days, followed by a ProK digestion then a standard kit or PCI cleanup... I get very high quality and purity DNA using this method.

If you have coral samples stored in ETOH, you will have to wash the sample thoroughly in distilled water. If you vortex your sample in distilled water, it also helps remove most of the mucus that the coral produces. Mucus can be a problem, especially as it can block silica based flow through tubes and can inhibit PCR.

Good luck!

BEAST software should be useful for you. It considers each gene information (seperate clock models, substituon model etc.) at the same time and it also handle with incongruence among gene trees. 

Please be aware that "sudo" means that the software has root = complete access to your computer and private data. It is a command for administrators and should be used only in such situations. 

I agree with Maria that the photomicrograph could be of  Trachelomonas armata(Ehrenberg) F.Stein 1878, with the following features:

I agree heavily with Jose Tomas Matus that using CDD to find families with matching domains is a very good idea. I would just caution that you may need to be careful about what you get from this search. If you pull a divergent protein where this domain is only a small part of the overall sequence you will wind up aligning only that domain and your tree will be a history of said region. This may not be a problem depending upon your circumstances, but it might be a major issue that gives you a false result.

One way of augmenting a CDD search is through a set of BLAST searches. I assume you already have accepted sequences for each of your families, plus your own sequences. Use a number of each from all three groups as seeds for BLAST searches and take a large number of hits (preferably, mostly those you know are not part of your groups). Ideally, these will coincide with your CDD searches and you can include a significant number of them to provide shape to your tree. A large and varied outgroup set can help ameliorate some of the problems of using a single too-distant or too-close taxon. If your new group falls in or sister to the two know groups with no outgroups among or between, then you have some good evidence that your 3rd group is closely related to the knowns.

STRUCTURE could be run with a large SNP dataset like you have but would not allow for phylogenetic reconstruction and while you could theoretically use *BEAST, it would likely be computationally intractable with that number of loci. There are, however, some other options in addition to what has been suggested above that may suit your needs. There is a likelihood-based inference method by Hearn et al. (2014) and also a quartet-based species tree inference method by Chifman & Kubatko (2014). I have successfully employed the allele frequency spectrum (AFS)-based inference tool ∂a∂i (Gutenkunst et al. 2009) to infer demographic history and phylogentic relationships among taxa using RNAseq data, and I would recommend giving this a try as well. Best of luck!

Sebastian and Xavier are right. Example: Sequences from two (haploid) populations

pop1:

ACAAAT

ACAAGT

pop2:

CCAAGT

CCAAGT

In this case, the internal variability Hs is 1/6=0.167 for pop1 and 0 for pop2. However, the average pairwise nucleotide diversity between the populations is 0.25 because of the fixed difference A/C between populations in the first base.

I have it if you still need it.

Check 1 ul of your DNA in an agarose gel (0.8%) made with TAE.

If you can not see  a high molecular weight band, change your DNA extraction kit. (or check with the company providing the kit).

DNA from members of the Fabaceae should extract nicely....If you perform your ITS PCRs with DNA from the stock, then something is going wrong with the extraction

Null alleles is a likely answer, and some samples will never amplify because of that (but as Roberta suggests, this might not be a major problem with analysis). If you decide to try one more amplification, you might try spiking the PCR of the reluctant samples with magnesium chloride. We discovered this helped some samples significantly - I would try increasing by just 1 mM and see if you get results... just be sure you are not getting non-specific priming... microsatellites are nice because you know where the alleles should be - if you get peaks/bands outside of that region with extra MgCl, back down on how much you are putting in...

Good luck!

Some people also like to digest again at a later stage to remove chimeric fragments that ligated together at the cut sites. 

Yes, it is H. norvegicus. In the case of potential neotypification these specimens should be of more interest than the previous as they are somewhat closer to the (potential) type locality. Although this is not too precise in Gunnerus' original description I would believe so. 

Cheers, Torkild

Dear Spencer, we have been doing some structure and admixture analysis on wild species populations, using also STRUCTURE and NEWHYBRIDS (among other approaches). At least in two cases (different species and scenarios), STRUCTURE identified, as different genetic clusters, two populations we suspect that correspond to a parental and an admixed population (in the absence of the second parental population). I say that we suspect that because we run the same pairs of populations in NEWHYBRIDS and we got  very good agreement with the two clusters identified in STRUCTURE and the two identified in NEWHYBRIDS. The interesting thing is that NEWHYBRIDS identified individuals from one of the populations as parental, and the individuals from the second as second generation hybrids. In both cases, we did not included (and were not identified) individuals from the second parental population. These results seem to make sense on the biogeographic context. In one case, we suppose that the second parental population is actually a source population from a founder event and that the admixed population is actually resulting from the admixture between propagule population and the native (pre-existing) parental population in the arrival place. On the other case, we have one population from one side of a mountain range and other in the mountain range (but not on the other side). We suspected that mountain population should be receiving genes from both sides and STRUCTURE identified it as an independent genetic cluster and NEWHYBRIDS identified it as an hybrid cluster.

So, trying to answer your question, I think that (in the absence of one of the parental populations), STRUCTURE is able to identify admixed population as a different cluster, but might provide no idea about the nature (admixed/no admixed) of that population. However, if you test the pair (putative parental + putative admixed) in NEWHYBRIDS, you might be able to detect if individuals are actually of admixed origin or not. I think this happens because NEWHYBRIDS algorithm explicitly searches for hybrid and parental classes and allways assume two parental classes (that it will try to reconstruct based on a bayesian algorithm). Even if you have no individuals from a second parental population, it might be able to reconstruct the hypothetic non-sampled parental population. At least with us, it seems to be doing that, at least in two different examples, and on a consistent way. Despite we only provide the genotypes to both programs, without any a priori information, it ends up consistently with the same outcomes.

From the scenarios I've already analysed with STRUCTURE, I found that when you have both parental and an admixed population, STRUCTURE detects the admixed population as a "gradient" of individuals with different proportions of genome assigned to each one of the two parental populations.

Well, I am not sure if this helps! If not, please apologise me for the long argument...

See the attached files.  It's one of several models  for rank abundance data against which data are often fit. Interpretation is complex, with arguments that provides evidence for neutral & niche based community structure. 

Hi Semir,

Thank you for your answer.

Since you have one species of Trichoderma only. The best marker will be SSRs. Don't do simple diversity but see the population structure which will help you in indentifying the various evolutionary forces shaping the fungus tragteroies

If you are having difficulty distinguishing them from the outset and have access to some soft tissue (liver is ideal), you can isolate mtDNA by differential centrifugation, then redesign primers to be specific to that sequence. As numts have a much lower mutation rate, and there is usually only one amplifying, they are usually easy to identify as a contaminating sequence. They are often useful in themselves as an outgroup for the real mtDNA- the low substitution rate means that you almost have the equivalent of a fossil ancestral genome to work with. So you might want to design primers for the numt itself too.

Thanks everybody! LogCombiner seems to work when I use an old version (1.7).

If you have not any information on your dataset, Evanno et al (2005) recommend to use an array of K-values = 1 to n+3, where n is the number of your local samples. I usually adopt this approach and I generally obtain a K-value << n+3.

I agree with Martin's comment, but I'd like to add that you should test the correctness of the concatenating procedure with the partition homogeneity test.

The size can vary.

But depending on the primer set used and the way you look for insert in your clone (PCR or restrict digest) you can have varied results.

It would also be interesting to note that you can pick up mitochondrial and plastid 16S genes, depending on your primer set.

However once you sequence the clones you can bioinformatically filter these or verify it.

Good luck.

Cheers

JS

About months ago I had the same question, so contacted the technical support. They say, that currently they do not support metagenimics data analyses.

Anyway, is you are going to do the workflow anyway, please do let me know! I have the same inputs (IonTorrent, 16s amplicons and workstation with CLC bio), but currently using QIIME to analyse data.