Hi I wanted to ask if it is possible to read a .nwk file with msprime or tskit

[AyomideOjo]

Hi I wanted to ask if it is possible to read a .nwk file with msprime or tskit

[benjeffery]

Hi @AyomideOjo, there are some methods in tsconvert that may be of use. https://github.com/tskit-dev/tsconvert

[AyomideOjo]

Good day, If you could please direct me to those fuctions that wouls would be amazing, because i am still having problems. Thank-you Ayomide Envoyé à partir de Outlook pour Android<https://aka.ms/AAb9ysg>

…

________________________________ From: Ben Jeffery ***@***.***> Sent: Thursday, August 3, 2023 5:14:23 AM To: tskit-dev/msprime ***@***.***> Cc: NotarealUsername ***@***.***>; Mention ***@***.***> Subject: Re: [tskit-dev/msprime] Hi I wanted to ask if it is possible to read a `.nwk` file with msprime or tskit (Discussion #2191) Hi @AyomideOjo<https://github.com/AyomideOjo>, there are some methods in tsconvert that may be of use. https://github.com/tskit-dev/tsconvert — Reply to this email directly, view it on GitHub<#2191 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AQAVJTMCQ5ZXYIKS6CQ7A53XTNMV7ANCNFSM6AAAAAA22SRLSI>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[AyomideOjo]

sorry I just clicked the link you sent, i found that after i made the post, but i am having some problems with working with self.kc_distance(tree) afterwards. thank you Ayomide Le jeu. 3 août 2023, à 05 h 14, Ben Jeffery ***@***.***> a écrit :

…

Hi @AyomideOjo <https://github.com/AyomideOjo>, there are some methods in tsconvert that may be of use. https://github.com/tskit-dev/tsconvert — Reply to this email directly, view it on GitHub <#2191 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQAVJTMCQ5ZXYIKS6CQ7A53XTNMV7ANCNFSM6AAAAAA22SRLSI> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[jeromekelleher]

Hi @AyomideOjo 👋

Perhaps you could give us a little background on what you're trying to do and we might be able to point you in the right direction?

[AyomideOjo]

Hello, I'm a McGill undergraduate working in a lab over the summer. I was asked to translate kc_distance from R to Python, where the tree data is provided from a .nwk file. It has been a surprisingly difficult task. Initially, I attempted to solve the problem using Multiprocessing and Cython; however, that approach didn't work out for trees of large sizes. Eventually, I found 'tskit' and was able to download it. However, the library function kc_distance only works with tskit.treesequence data.

I then tried converting the Newick trees to tskit.treesequence trees, but I encountered some problems. Firstly, the data I received had branch lengths of 0.0, which the from_newick() function rejected. To resolve this, I changed the min_edge_length variable to be equal to sys.float_info.epsilon * 1.25.

Another error I stumbled upon was that the tskit.treesequence tree generated by the function seems to have nodes labeled with IDs given by the function add_row. Furthermore, the function kc_distance appears to call another function named check_kc_distance_samples_inputs to verify that the tree's tips have the same names. However, my error here is that the samples attribute returns a list of IDs, which do not match the names of the tips in the data I was provided with. I presume that the add_row function assigns IDs incrementally, starting with 0, then 1, then 2, and so on. Consequently, the for loop in the function check_kc_distance_samples_inputs detects that two leaf nodes have different IDs, leading to the error TSK_ERR_SAMPLES_NOT_EQUAL. To address this, I implemented my own function to change the IDs so that they are equivalent to the tip names in the data. Unfortunately, the IDs are immutable and cannot be changed.

Currently, I am attempting to create nodes myself, which I have managed to do. However, I am struggling to find a way to add those nodes to the tree. If you have a solution or potential remedy to this problem, I would love to know.

Here is my current directory where I am working: https://github.com/AyomideOjo/grasses/tree/main/Final/Enoch_O.py
My other versions are here: https://github.com/AyomideOjo/grasses/blob/main/Ayomide_O.py
and here: https://github.com/AyomideOjo/grasses/blob/main/Parham_G.py
My functional but slow cython file:https://github.com/AyomideOjo/grasses/blob/main/cython_code.pyx

Thank you for any help you may be able to offer
Ayomide Enoch Ojo

[jeromekelleher]

I see - unless you're specifically interested in using tskit I think you'd be better off using dendropy to read in the newick trees. You should be able to write a KC distance function using dendropy without too much difficulty. Tskit isn't really set-up for general phylogenetics work based on an input newick tree right now.

Using tskit would likely be faster all right though, if that's an issue.

[AyomideOjo]

Alright thank you, I gave the documentation a read its fantastic, thank you, seems like exactly what ive been looking for.