Phylogenetic Tree - Science topic

Outgroups are from taxa that have been chosen apriori to analysis because they are well know to be outside of the group of interest. One cannot infer the root of the tree without an outgroup to anchor the root. So calling this clade an outgroup would be circular and arbitrary. Indeed, if the support for the apparent clade is small than you really don't know the topology of the tree either. That is fine but unrooting the tree altogether would give you an illustration of what you do know. Present it as an unrooted tree until you have independent data to define an outgroup.

@James Bonacum, what you have just written is highly informative. Thank you very much

If you have the complete mitochondrial genomes aligned, you can use the exclude command:

And there are other ways to create a "mask" sequence in a multiple sequence alignment file, and then take columns from your alignment that are not masked.

It is always best if you use the identical region from each species in the alignment. Ideally it would use the full 16S gene and they would be almost identical in length. Because each species has multiple copies of the 16S gene, you can get different genomic regions. They should be identical or almost identical in sequence. Thus, if you aligned them you in a tree you should get a very tight cluster for each read from the same species and the tree structure would come from the different species. Alternatively, try building a consensus or just alignment from a single species to check that they are almost identical.

Just as Jens Rohwer said, first, align each marker separately then you will need to concatenate the sequences to form a single "super-alignment" file. this will serve as the new input sequence for your phylogenetic tree.

It depends on the purpose. I'm a bacterial taxonomist and we use core genome based trees in our studies. The two pipelines I'm familiar with are UBCG and BPGA. In short, UBCG creates a tree based on concatenated sequences of 92 conserved core genes while BPGA defines a set of core genes shared by all the genomes.

329 / 5.000

The clades are on the right, and the left is the primary origin of the genus. The numbers are the distances in relation to the origin; the further away they are, the more distant they are, the more likely there was deletion, mutation, or evolution. In mega11, the color in the yellow case, or orange, is the color of interest. It groups the branches by species. In this case, the numbers are measures of genetic distance between the DNA sequences of the samples, that is, they are evolutionary changes that give us a genetic similarity. Please, in mega11, compare the sample of your interest with the other closest species and you will see if there are samples preserved or not by asterisk. At your service.

Seaview (https://mybiosoftware.com/tag/seaview) is a reliable and easy to use program to learn phylogenies. It is a compendium of really excellent independent program that accomplishes alignment (use the Muscle option) and really fine distance, likelihood and parsimony inference programs.

Similarly you should look at TreeViewer also with a nice graphical interface (https://treeviewer.org).

Other excellent programs are for likelihood:PAUP, Raxml, IQ and for Bayesian: MrBayes. One of my favorites is SplitTree for tree diagnosis. And for publication quality trees, try FigTree. These last few are my go-to programs but may take a bit more training - which you should get anyway.

These all have very good manuals. Also I will be making my many years of lectures available soon on my researchgate page. Will take questions.

Without seeing you multiple sequence alignment, we can only guess. But most likely there were many other sequences with greater identity, like 99.95 to 99.99% identity.

I was built using two distinct phylogenetic trees created by MEGA X. The first tree was created using the ITS sequences of 24 different species of mushrooms, and the second tree was created using nuc-LSU sequences, which contain the same number of sequences (taxon), using the calculus models of Tamura-Nei and Maximum likehood

Can someone help me to discuss the result (tree)?

EMBOSS https://emboss.sourceforge.net is an open source software collection that contains a wide variety of sequence utilities. Its utilities are also available through the web: see https://emboss.sourceforge.net/servers/ for a list of servers.

You want to reverse complement https://www.bioinformatics.nl/cgi-bin/emboss/revseq and translate https://www.ebi.ac.uk/jdispatcher/st/emboss_transeq your sequence

(Of course, you'll find both utilities on the same server, but I just gave you the links to the first hits I got when googling for "emboss complement" and "emboss complement"

Hi,

As David L Remington said, those of us who are not experts on diving beetles can only speculate without seeing the actual data. The insects have been evolving on earth for quite a bit longer than the mammals have, and it seems possible that diving beetles could have a very distant common ancestor, or perhaps even convergently evolved a few times from non-diving beetles. Complete mitochindrial genomes are often very accurate at phylogenetic reconstruction in comparison to fossil record etc. But COI gene (or protein) alone is often a bit misleading. Also, mitochondria can sometimes tell a slightly different story than nuclear chromosomal genes.

You must also beware, that if you search GenBank and use a lot of data from there, that it is not uncommon for samples to be mislabeled for genus/species. This can happen either because the authors mis-identified the sample, or because there was some type of PCR contamination event or mis-labeling or material in between the sample and the GenBank entry.

Hello, Stavros. You will need to open and edit the tree in a vector graphics editor, such as Inkscape (free software), Adobe Illustrator or Corel Draw.

Most folks would concatenate, align, & trim off any sequence from the ends that are an overhang. Then build a tree.

Good luck!

Thanks a lot Mam

Inferring phylogenetic trees depends on nucleotide sequences. When the sequence fragments are partially missing, MEGA's "Align" function is good at aligning the valid parts to build the phylogenetic tree. However, I don't quite understand what you mean by constructing a phylogenetic tree using only biological names. If I guess correctly, are you trying to say that the outgroups of the phylogenetic tree are constructed by relying only on the biological names without any sequences? It seems to me that this can't be done in MEGA, first of all the developmental tree needs to be constructed based on the differences or affinities of the sequence fragments, and this obviously won't work if there is no information about the sequences. However, you can "artificially" add outgroups by manually modifying the tree file, but the root of the tree, must be considered before adding outgroups. If this is not taken into account, then the tree is probably incorrect. In addition, when you choose to add outgroups manually, rather than relying on a program to calculate them, you may need to ignore the branch lengths, since it is obviously impossible to know the length of the outgroups in the tree in this situation. Overall, adding outgroups manually is not recommended.

It depends on your goals. After you obtained the consensus sequence (from forward and reverse), you can perform Blast analysis to find the most closely related sequences from other strains/species. Download the number of sequences you find useful to your analysis (to be included in the tree), align them using for example MEGA as proposed by Ramesh, and build also the tree in MEGA. Write me if you need any additional help or guidance.

Reference sequences should be from type specimens such as holotype or paratype or epitype reference sequences. GenBank is full or erroneously labeled (unvalidated) sequences. There is no threshold (percent similarity based on BLASTn searches), it is either a type specimen reference or not. When you BLAST your sequence you can select type specimens, which will only return type specimen reference sequences. You may also need to peruse the literature to locate type specimens. Most articles place type sequences in bold or some other notation to indicate type (validated) sequences in a table or phylogenetic tree. Hope this helps!

If I'm not mistaken, the taxonomy of archaea falls under the International Code of Nomenclature of Prokaryotes, together with bacteria. As a bacterial taxonomist, I use the LPSN database as a reference (link below). Taxonomy is a turbulent field, where literature can get outdated pretty quickly. Up-to-date curated databases are a good source for current taxonomy and nomenclature and can be cited in scientific literature.

I recommend reading a bit about how the taxonomy and nomenclature of prokaryotes actually work. You can check out the Nomenclature and FAQ pages on the LPSN website.

On the website linked below, you can find the names of archaeal phyla, including many with Candidatus status. The database also links to many useful resources, such as the effective publication of the taxon you're currently viewing, synonyms, and other notes.

Actually, the two domains and seven kingdoms of prokaryotes were officially included in the code one month ago (link to publication below), before that, domains and kingdoms were not included in the code at all. The same happened with Phyla a couple of years ago, before that, we only had species, genera, families, orders and classes. I guess the linked paper on its own would answer your question, but I think the database comes in handy when you want to browse the ranks.

Yes you can, but do not remove gaps that are generated after alignment within the aligned sequences.

There might not be a way to increase the bootstrap value without adding other sequence regions. Assuming your sequences are aligned correctly, poor bootstrap support means there simply aren't enough informative variable sites to support one branching pattern over another.

You don't say which tree construction method you used; e.g. neighbor-joining vs. maximum likelihood. Sometimes different tree construction methods will provide a little better resolution than others, but each method can also introduce its own biases, so it's not a good idea to go "shopping" for the strongest bootstrap values.

these codes are IUPAC codes for mixed bases so W is A or T, R is A or G and y is C or T. Polymorphic bases are common and the individual bases will form recognisable haplotypes . You cannot remove these bases as the comparison software will see base deletions which is not true but all phylogenic software will recognise these codes and deal with them appropriately

Bayesian Inference method in MrBays or Maximum likelihood method in RaXML or IQTree methods can be used.

IQTree offers an online platform to construct phylogenetic trees. You can access to that service via following link.

It looks like your Crytosporidium isolate is misidentified, and not truly a member of that Genus. But it could also be that your sequences has some region misaligned, to create such a large distance from the other sequences.

There are a few stats you might report. You might look at length in bp, or number of parsimony-informative sites, or number of variable sites, or some other metric. I normally use this summary tool https://github.com/marekborowiec/AMAS

Quite possibly, that is called species with 'deep split' barcode divergence. We presented many examples in our North American Noctuoidea barcode library a single species with 18 BINs! First, you need to make sure there's no misID involved, very common mistake. Could be an overlooked species despite having the same morphology and genitalic structures in adults, but different ecology and behaviour, pheromones, spatial distributions, etc.

You can also give it a try with the MEGA Platform.

Home (megasoftware.net)

There are many ways to do it. But for almost all of them, the sequence names will have to be created with labels that will then be used to color each OTU by some aspect of the name.

The Interactive Tree of Life project is one good resource:

The HIV Databases at LANL provides another simple tree coloring service:

The Augur and Auspice tools are also very helpful:

In Figtree, you can select individual OTUs, groups of OTUs based on a character set in the names, or select clades, and then color the brnaches, the OTU names, or both.

Dear Ulyana Simakova

Here is a great tutorial (with references) for estimating species trees with *BEAST: https://taming-the-beast.org/tutorials/StarBeast-Tutorial/

Hongzhuang Chen I am afraid that you can lose a lot of information from such comparison. But, it can be applied (and very useful) to illustrate the differences supported statistically by analysis of the original data (sequences).

You may find it on the NCBI database if you search for dengue sequences like this

Maybe it is not an error. Sometimes samples from different regions are closely related to other region plant samples because of a lack of genetic structure (best case scenario) or a lack of resolution of the markers you're using (worst case scenario). Do you have any literature to support a hypothesis of them being differentiated?

Give it a try with different phylogenetic software, although I would assume that might not be the problem.

Good luck!

In addition to the comment done by Susanta Roy , I would like to share with you this paper (Bast 2013 Nature Protocol Exchange) that describe the generic sequence analysis protocol using software like MEGA. Prior to the phylogenetic tree reconstruction, you can use Bioedit to Edit DNA sequences.

Thanks for the clarification. So adding the "the bootstrap has an element of randomness " Note, will not affect the answer to the question regarding the number of well-supported nodes. Is this right?

As you present your tree, there is 99% percent support. No question.

The goal of the tree building is to distinguish homologous sites that join organisms from those the split them.

So intuitively you know that sites that have only one or two sequences cannot be phylogenetically informative. Hence you might as well trim the alignment at the point that no phylogenetic data.

On the other hand, a likelihood method, for example calculates the likelihoods of each site in the alignment and then sums the site likelihoods to give the likelihood score of the inferred tree. Any site that does not have much information will necessarily have a low likelihood score so will not contribute very much to the overall score so, in a sense, these sites have little influence on the likelihood tree.

There is a large literature that describes the pitfalls of uneven lengths. I tend to think that leaving some sparse sites in the alignment will at least contribute the likelihood of any phylogenetic patterns among the taxa that are included. This is arguable, however.

Haven't tried it myself yet but R package "apex" seems to do just that.

As a general rule, use maximum likelihood for your final tree. You may get a sense of the relationship with a preliminary neighbor joining tree but the tree will possibly differ if systematic bias is present. ML will correct that if the model of sequence evolution is the right one. A preliminary quick NJ tree is very useful for checking the quality of your data and its alignment. Any very very long branches suggest a good look at the data and the alignment to look for errors.

Let me know if you need help. Would be good to know more about your data set. How long are the sequences and how many sequences? What kind of phylogenetic software do you have access to?

Hi Cristian,

I made some trees using the HKY+G model in PhyML (part of the free Seaview package) using your fasta alignment. Both are presented with the branches proportional to the ML estatimate. The first uses all taxa but the branch to the two apparent outgroups is so long that it makes the rest of the tree less readable. These two are clearly very different from all the others. The second tree leaves them out for clarity.

The tree that you included is called a cladogram because the branch lengths are arbitrary. The idea is just to get a notion of the branching pattern. You can see that a cladogram can be very misleading.

These trees show the branching pattern and the branch lengths which are proportional to the number of substitutions on each branch. This so-called phylogram gives a clear picture of the rate of evolution on each part of the tree.

Let me know if you have any questions at all.

I had the same problem with MegaX, try running it with lower threads. Was able to construct a ML tree using only 6 threads (My laptop has 12 threads).

Try IQ-Tree server once. If there is any problem send me the alignment file, I will try once.

Depends if you want to use long or short reads. There are facilities like the Sanger Institute that do the sequencing. As for the assembly of the reference genome, there are pipelines (eg. ). Usually you need about 3 micrograms of high quality DNA. Your model being an insect should not change anything once you have performed the DNA extraction step. Maybe ask around your own university or nearby ones if any of the labs have an Illumina sequencing platform.

Phylogenetic tree construction can be a complex process, and there are several factors to consider when interpreting the results. The choice of statistical method (such as maximum likelihood or Bayesian) can influence the topology of the resulting tree, as can the choice of evolutionary model and the parameters used in the analysis.

One important statistic to consider when evaluating the quality of a phylogenetic tree is the bootstrap support value. This value represents the percentage of times that a particular branch is supported by resampling the data. Generally, a bootstrap support value of 70% or higher is considered to be significant support for a particular branch. You can also evaluate the posterior probabilities in a Bayesian analysis, which provide a measure of support for each branch.

The workflow I usually take is to build several trees with a low number of iterations or bootstrapping (to save time) from the same data input using different models. By comparing their topology, I get a reasonable insight into how robust the tree is. When almost all models result in a tree with similar topology, you can choose one of these methods that seems appropriate and rerun that method with more bootstraps or iterations for statistical support. One important factor to consider is the overall topology of the tree. If the tree has good support values for all major branches and is consistent with previous research on the organism(s) in question, it is likely a good representation of the data.

Another thing you could do if you have constructed multiple trees using different methods or parameters is to compare the trees using statistical tests such as the Shimodaira-Hasegawa test or the Kishino-Hasegawa test. These tests can help you determine whether one tree is significantly better than another based on the likelihood scores and the number of parameters used.

Finally, it can be useful to visually inspect the trees to look for patterns and outliers and compare this with the alignment. If there are a few sequences that seem to be clustering in unexpected ways or are not well-supported by the data, it may be worth investigating further to see if there are any errors or issues with the data. You can choose to remove an outlier to reduce its interference with correct tree building, and/or you can add other sequences which allows for clustering when a branch in the tree does not have many closely related species. A lack of sister species can result in a branch that starts apparing in unexpected places as it does not really have a clustered place in your tree.

I hope this helps! It usally takes quite some work to construct a good tree but eventually you will get there.

Hi,

It depends on the type of organism you are interested in. The Genus and Species names for mammals, insects, shellfish, freshwater fish, ocean fish, bacteria, viruses, fungi, etc. are not all equally accurate. For many organisms, the genetic relationships do not match up well with other classification and nomenclature systems.

Phylogenetics can be very helpful in sorting out genetic relationships, but one must be careful to understand that "gene trees" or phylogenies based on one or a small number of genes do not always agree with "species" or populations trees based on larger data sets, due primarily to horizontal gene transfer. The rates of horizontal gene transfer vary greatly between types of organisms.

Aloha Sohail, I can say a few words here. First point: transition mutations are generated at higher frequency than transversions.

Transitions (tn) are substitutions of purine for purines or pyrimidines for pyrimidines, transversions (tv) are purines for pyrimidines and vice versa, these are more rare in coding sequences, tn tend to have a diminished phenotypic impact relative to tv, along the lines of the way that the third position substitutions in a reading frame are more common, and tend to be evolutionarily, and phenotypically, neutral, in this case due to third position wobble of course. Thus 1st and 2nd position substitutions are scored as more important (rare) in phylogenetic data, again akin to tv/tn ratio.

You're welcome Omkar Damle

Hi,

I don't know much about flatworms, but for a great many types of organisms, including invertebrates and vertebrates, the mitochondrial genomes and even more often mitochondrial genes such as nadh and cox-1 are studied for phylogenetic classification and taxonomy.nomenclature. I searched GenBank for Macrogyrodactylus and only found one mitochondrial genome. I then used BLAST with the cox-1 gene from that genome and again found only one Macrogyrodactylus plus a few related flatworms. I aligned the cox-1 genes and have attached the alignment and tree here.

As for outgroups, in general you should select the next closest relative of the group you are interested in, and not something very distant like mouse or insect, for many reasons. Even within the flatworms, the 0.5 scale bar in my attached tree indicates a huge distance between these organisms. You can see that within a species (G. lucii, G. botnicus) there are variable degrees of diversity which is probably more of a nomenclature issue than a biological issue.

It depends what do you mean by long branches. You should attach an image and point them out. Sometimes Bayesian methods are known overestimate branch lengths.

Best,

Denis

It is possible for sister taxa to have the same DNA sequence, but it is rare. Sister taxa are defined as two taxa that share a recent common ancestor that is not shared by any other taxa. This means that they are expected to have some similarities in their DNA sequences due to their shared ancestry, but also some differences due to the accumulation of mutations over time.

However, there are some circumstances in which sister taxa may have identical or nearly identical DNA sequences. One possibility is if they have recently diverged, and there has not been enough time for mutations to accumulate in their DNA. Another possibility is if they have undergone convergent evolution, where two distantly related lineages evolve similar traits independently. This could result in the same or very similar DNA sequences evolving in two different lineages.

Overall, while it is rare for sister taxa to have identical DNA sequences, it is possible under certain

Including an outgroup in phylogenetic tree construction is important because it helps to root the tree and determine the direction of evolution. The outgroup is chosen based on its evolutionary distance from the ingroup, meaning it is a taxon that is closely related to the ingroup but not a member of it. The inclusion of the outgroup can affect the placement of the other taxa in the tree, including its position within or outside of the ingroup.

The difference in results between the two methods you used, neighborhood joining and maximum likelihood, could be due to the different algorithms and assumptions used in each method. It is not uncommon for the placement of the outgroup to affect the position of the ingroup in the tree, and the results from each method should be evaluated and compared to each other and to the available literature to determine the most reasonable placement of the taxa.

If you are concerned about the placement of the outgroup and its effect on the results, you could try using a different outgroup that is more distantly related to the ingroup or choose a different method that may be less affected by the outgroup. However, it is important to keep in mind that the choice of outgroup and method should be based on sound evolutionary and statistical principles.

You need at least representative sequences for each OTU.However, if you are working with ITS sequences better forget about it. Unless you are working with a single genus, the ITS is useless for phylogeny due to lack of conservancy. Fine as a barcoding sequence, but not really suited for phylogenetics in general.

Although people who answer to the question interpreted the question and answered it accordingly, however, the way this question was asked is wrong.

Hi, I meet the same problem when I build a NJ tree, I would like to know what this means and how to fix it.

Overall, since you are interested in the relationships among the members of the ingroup (the outgroup is assumed to be outside), then you are better off with missing data there than you are if you have a lot of missing data among the ingroup taxa. The really tricky problem is when you have complementary patterns of missing data, such as

agctagct--------

--------agctagct

Many of the phylogenetic programs will view these two taxa as identical, because missing character states are treated as a wild card that is considered to be the same as the character state of a taxon it is compared to. This is definitely something to avoid. See Cladistics 34:467 (2018) for an example.

Hi Daniel,

When using secondary calibration points, it's highly recommended to use a few. But when you want to incorporate fossils or biogeographics events you should use as many as you can. This because when you implement secondary calibrations it is highly probable to incorporate errors and bias associated to the methodology used with the secondary calibration.

Here at the Institute of Biology-UNAM, there is a very good class on molecular systematics and a module focuses on dating analysis including calibrations. It is offered by Dr. Susana Magallón. I include some of his references in case they help you.

Hope this will help you!!!!

Saludos.

So you are short three trees. You must have another three ways to root the tree.

Hi Teresa. Did you check if the weird sequence is in the right orientation? i.e. maybe you need to reverse-complement it. Good luck!

What are you using to plot with? You need to plot your tree as a cladogram rather than a phylogram. Both Dendroscope and Figtree have options for this (either explicitly setting plot type as cladogram, or something like "use equal branch lengths"), or if plotting in R using ape:::plot.phylo you can set use_edge_length=FALSE

The publication supplementary materials might have the tree file. Or, you could contact the authors and request the .tre or nexus file.

Thanks.@

CMW is the institution code (https://www.ncbi.nlm.nih.gov/biocollections?term=CMW%5BUnique%20institution%20code%5D) that isolated the strain and 14799 is the internal strain number. Depending on the type of sequencing data the authors may decide to store it in different repositories. For example, Genbank does not have an entry for that strain (https://www.ncbi.nlm.nih.gov/bioproject/?term=txid5158[Organism:exp]). I advise using the NCBI taxonomy browser to find related records for a specific strain (https://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?lvl=0&id=5158).

Results of Blastn and phylogentic analysis can vary significantly depending on the sequence quality and size of contig. To actually delineate a bacterial species its always recommended to use polyphasic (i.e., genomic, chemo taxonomic, morphological parameters) approach. In your case it is difficult to decipher whether you have used 16S sequence extracted from whole genome or did a Sanger sequencing using specific primers for your analysis.

However, you can still try to use EZtaxon tool (https://www.ezbiocloud.net/) to compare identity of your Blastn results and to also check if there is any contamination in your contig.

Gosiame Boitumelo Malepe A phylogenetic tree is constructed in four separate steps: (Step 1) find and get a collection of homologous DNA or protein sequences, (Step 2) align those sequences, (Step 3) estimate a tree from the aligned sequences, and (Step 4) show that tree in such a way that the important information is clearly conveyed to others.

Try TASSEL software to convert your SSR data to newick file using NJ or UPGMA and then use MEGA software to get your phylogenetic tree.

Best !!

Dear Researchers

While I am reconstructing the phylogenetic tree, during the alignment step, I am getting an error message "Application error finalizing MUSCLE alignment: Unable to open file". How can I resolve this issue?

Thank you.

How to edit the scale of a MEGA X phylogenetic tree?

Thank you all for your replies.

Good day

It depends on what the purpose of the tree is. Maximum likelihood and Bayesian methods require a lot of computation, and it may takes weeks to get the result. Neighbor-joining methods are far quicker and tend to give the same tree topology but perhaps with less accuracy of the branch lengths. For most data sets, the accuracy of the alignment, and the type of data uses (DNA vs amino acid, single gene vs multiple genes, nuclear genes vs mitochondrial, etc) is far more critical than the type phylogenetic analysis software used. Also, choosing the model of evolution and a few other factors are more important that the phylogeny method.

It in edit text

It really depends on what genus of bacteria you are working with, and what level of phylogenetic resolution you want. Some bacteria such as Shigella, Escherichia, Salmonella have been extensively studied down to individual clonal outbreak lineages, while other bacteria are mostly represented by uncultured environmental samples with no information about them. The rules for taxonomy and nomenclature of bacteria are also sometimes seemingly confusing. For one example a bacteria that produces botulinum toxin and is isolated and characterized after killing one or more humans will be named as "Clostridium botulinum" while the very same bacteria found in the environment or after killing a duck or any other non-human vertebrate could be named "Clostridium butyricum" or "Clostridium sporogenes".

@Shin Murakami really thanks for your recommendation, it made me a further understandind of the basic rules of PSI-BLAST~

Rishima Maheendran , the procedure is to first get the sequence of these genes and align them, there are many alignment software u can use, such as ClustalW or Muscle etc. After, you plot the tree, MEGA software should do. For gene families, they are probably going to be paralogs but drawing a phylogenetic tree would allow you to see the relationship between them

Kenneth Joseph Chan Bureros Just try to cluster all sequences from the same organism at 100% identity using cd-hit. If all sequences are identical or <1% distance or any negligible distance then you don't have to worry and you can pick any one sequence.

The fact that Ajay Saini mentioned is true and 16S rRNA being very conserved in the same organism will always differ more than in other organisms. Hence, after picking anyone sequence and doing Multiple Sequence Alignment, you will observe some of the nucleotide positions you have to clean.

So in my experience, it doesn't make any huge difference in the final phylogenetic tree. You can always try with different sequences and play to choose the final sequences and tree.

Is this from a single sequence/region, or is it from a concatenation of multi-locus?

IQ-Tree automatically estimates the substitution model that best fits your data (the one with the lower BIC value, in this case).

Imagine bootstrap values as saying that n out of 100 runs, that was the obtained topology.

Branch length is relative to what you're analyzing, and unless you're estimating divergence times, I would probably not get too deep into it (although I might be wrong).

Probably this paper will help you clarify the mind: https://academic.oup.com/sysbio/article/48/4/814/1627565

Cheers!