I think your best option might be Phylip for this, if I remember correctly there are several programs that can input a distance matrix. You just need to select the one for your input type (see: http://evolution.genetics.washington.edu/phylip/doc/distance.html). 

With R, you can easily generate a dendrogram using a distance matrix as the input. Two R packages that can do that are "hclust" for hierarchical clustering and "ape" for neighbor joining methods.

An example of hierarchical clustering is here: 

http://stackoverflow.com/questions/21442624/drawing-dendrogram-from-pre-calculated-distance-matrix

and this for the neighbor joining method:

http://r-eco-evo.blogspot.com.es/2007/09/neighbor-joining-tree-with-ape.html

If you are not familiar with R, I would recommend you to use Rstudio, which is a very user-friendly interface for R. https://www.rstudio.com/

good luck!

In a strict sense, the distance-based tree should not be called phylogenetic tree. Rather, it is a dendrogram which is clustered by calculating similarity.

Yunpeng Zhao is quite correct.  You cannot generate a phylogenetic tree from distance matrices although there are many programs that will generate a dendrogram from such data.  Phylogenetic trees are based on synapomorphies, which are characters.  You have lost all character data by reducing the taxon X character matrix to a taxon X taxon distance matrix.

Aside from arguments about what methods are worthy of the term "phylogenetic," the more important question is what the distance matrix is based on.  If it is DNA sequence data with appropriate nucleotide substitution models used to estimate numbers of changes from the percent dissimilarity, and the common ancestries are recent enough, there can certainly be good information to infer phylogenies using such methods as neighbor-joining.  To be sure, all the site-specific information is discarded in the process, though.  However, if these are merely some form of morphometric distances, attempting to make phylogenetic inferences becomes highly questionable.

If you are working with DNA or protein sequences, keep in mind the reason why alignment is important.  The goal is to have homologous sites (originating from a common ancestral site) aligned together so that the evolutionary history of successive divergences from the common ancestors can be inferred.  This is essential regardless of what methods you are using to infer phylogenies.  I know there are methods for combining the alignment and phylogenetic inference into a single process, but I don't have experience with these so won't comment on their value.  To use them, however, you would still need the original sequences.

Thank you all its appreciating! 

@Yunpeng Zhao, @Donald G. Buth, Exactly I am completely agree with you on this point, But if we have novel species and cant determine its morphology then DNA based similarity is the widely used method these days to assign phylogenetic relationship. But it might be possible what we predict may not be exact what it should be but still we found somewhat similar closely related group of species on which we can assign that novel species and it may be more accurate than morphological based method.

David Reminton has provided a comprehensive answer.  I would endorse it.  In addition, I would suggest a very user-friendly and free program called MEGA (http://megasoftware.net/) for inferring your phylogeny.  You can input distance matrices directly into the program.

In MEGA6 I tried but it didnt take distance matrix, by the way I have solved the issue with AAF (Alignment Assembly Free) phylogenetic tool, and now I got both the results including distance matrix as well as dendogram tree.

Once again thanks for your answers, It helps!

Good to know that you have your tree.  But MEGA does accept distance matrices - I use them all the time.  It is true that the matrix needs to be formatted in a certain way for MEGA to read it.  See attached file for the format.  MEGA will accept it once it is formatted this way.  One of the many reasons I find MEGA useful is its excellent tools for manipulating the tree in the "Tree Explorer".

Well I found it useful thanks @Sudhindra.

But can you tell me how to generate that format. is there any tool or have to generate manually from standard distance matrix file?

Actually, I am working on a tool that will do that (as a side utility) in addition to other things, but a release will take a bit more time.  At the moment, I'm afraid you'll need to do it manually.  MEGA will convert fasta and other sequence formats to the MEGA format, but will not convert a simple triangular distance matrix to the MEGA format - at least not in MEGA6.  I haven't checked MEGA7, which has recently been released.

If you read my question, there is clearly mentioned the tool with same functionality, never mind, Below is the tool link:

AAF [Alignment and Assembly Free Phylogenetic Tree]

http://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-015-1647-5

Dear Jan

Thanks Jan, I m sorry for misunderstanding, it was happened just because you added that line ">It will be appreciating if someone suggest any tool which can generate phylogenetic tree without alignment". So I thought you were asking the same question.

Thanks for your suggestion.

AAF uses assembled scaffolds or reference available genomes as a input to generate phylogenetic tree.

Dear Hemant,

You can use R program which has sophisticated libraries for creating distance based phylogenetic trees from input distance matrix. If provide tools to create different phylogenetic tree construction methods like UPGMA, NJ, UNJ, BONJ and minimum evolution, etc.

you will need libraries like ape, stats, cluster, phangorn, ips

http://www.phylogeny.fr provide a multiples ways to generate a phylogenetic tree, including BIONJ which can use a distance matrix.

You can try these links below

http://etetoolkit.org/treeview/ http://www.trex.uqam.ca/index.php?action=newick http://www.trex.uqam.ca/index.php?action=trex http://www.trex.uqam.ca/index.php?action=phylip&app=fitch http://itol.embl.de

Hello,

I also want to calculate a NJ tree out of my SSR genetic distance matrix. Therefore, I tried to create a file like the helpful example of @Sudhindra R Gadagkar, but Mega tells me that there is an ; instead of a distance. Has anyone suggestions to solve this problem? The file is in the attachments

Thank you

I had no problem in using Mega to obtain neighbor-joining trees from the distance files supplied by Julia and Hemant (I just had to replace the commas with points). However, such trees lack bootstrap values.

Hello

I am studying genetic diversity among in populations I tried many times to draw the tree but it seems difficut I have 4 population and the FST matrix but nothing works for me Mega is very complicated especially when I try to convert from Txt to Mega file.

I did the genotyping of STRs

Thank you

@Safia, please use the example attached (DistanceData.meg, which you can open in a text editor like Notepad in Windows) to manually convert your text file to Mega format in a text editor. I have left all the parts that should be specific to your data in UPPERCASE. So change them accordingly (and also the values in the distance matrix, of course). Leave everything else as is. Believe me, it is straightforward.

Name your file with a .meg extension, and assuming you have installed MEGA, it should open in MEGA if you double-click it. Go straight to the Phylogeny tab and do your phylogenetic analysis. For this example, I ran an NJ tree. Also attached.

Good luck!

Sudhindra R Gadagkar Thank you very much I will try it and tell you the results

many thanks

Sudhindra R Gadagkar

same issue the program asked me to convert the file and it doesn't work even I tried to open the file you uploaded

Sudhindra R Gadagkar I delete Mega and downlooad it again , and this is the same error message

Safia A. Messaoudi I checked the file Sudhindra R Gadagkar sent and i believe it has blank spaces after the matrix values, maybe it lost proper format during transfer.

In any case, I do not know hot to handle Mega. However, I just plot a huge NJ tree from incomplete (lower triangle) distance matrix using the good and old PhyD (available at http://www.atgc-montpellier.fr/phyd/usersguide.php). It's straightforward. Hope it helps.

Filipe Pereira Matteoli thank you

i will

Sudhindra R Gadagkar

Thank you mate

I am checking these messages after a very long time and seem to have missed the posts of Safia A. Messaoudi where they mentioned a problem with Mega in using a distance matrix for phylogenetic inference. I hope it has been resolved. In the meantime, I thought I'd mention that I recently published a computer program called PHYLOM, where I convert a matrix of measurement data (say metabolic test data for m tests and n species) into a pairwise distance matrix and then use Neighbor-joining for phylogenetic inference. Bootstrapping and rooting are available options. The paper is freely available here: Article PhyloM: A Computer Program for Phylogenetic Inference from M...

try this perhaps?

http://www.trex.uqam.ca/index.php?action=trex&menuD=1&method=2