search

veg / hyphy

HyPhy: Hypothesis testing using Phylogenies
http://www.hyphy.org
Other
219 stars 69 forks source link

interpreting aBSREL results #391

Closed vabadu closed 5 years ago

vabadu commented 8 years ago

Hello Developers,

I have run aBSREL on a number of data sets from the website: http://test.datamonkey.org/analyses. The front page of the results have always reported that 0 branches were found to have experienced episodic diversifying selection. This is true even for data for which the Paml branch-site test of positive selection and Hyphy BSREL have significantly detected a branch with positive selection. I notice in the aBSREL results, under the "Table" tab, the message after the name of each branch of "(not tested)". "Not Tested" is also noted for all branches under the LRT column in the table. This makes me wonder if I have not run the program correctly.

Your time and consideration of my question is appreciated, Vanessa

spond commented 8 years ago

Dear @vabadu,

Can you please supply a link to an example aBSREL results page where this behavior is exhibited?

Sergei

vabadu commented 8 years ago

Here you go: http://test.datamonkey.org/absrel/56c383d39dd92ef47c708fe9

spond commented 8 years ago

Dear @vabadu,

You are correct: something is definitely not right here, because at least three branches (with point estimates of ω2 > 1) should have been tested for selection. I'll ask @stevenweaver to look into it.

Best, Sergei

vabadu commented 8 years ago

Thank you so much for your help. I look forward to hearing back from you.

Vanessa

From: Sergei Pond notifications@github.com Sent: Tuesday, February 16, 2016 2:51 PM To: veg/hyphy Cc: Vanessa Lamae Bauer DuMont Subject: Re: [hyphy] interpreting aBSREL results (#391)

Dear @vabaduhttps://github.com/vabadu,

You are correct: something is definitely not right here, because at least three branches (with point estimates of ?2 > 1) should have been tested for selection. I'll ask @stevenweaverhttps://github.com/stevenweaver to look into it.

Best, Sergei

Reply to this email directly or view it on GitHubhttps://github.com/veg/hyphy/issues/391#issuecomment-184889791.

spond commented 8 years ago

Dear @vabadu,

Just to confirm, the equivalent analysis executed in HyPhy (https://github.com/veg/hyphy-tutorials/tree/master/selection#use-absrel-to-find-lineages-which-have-experienced-episodic-diversification) yields the following output, showing 3 branches under selection.

...
Summary of branches under episodic selection (7 were tested, of which 4 required optimizations) :
    Node4 p = 4.993840995837218e-06
    WHITESOUTHERNRHINO p = 0.007700264980029792
    DONKEY_TRIM25_1 p = 0.03754281322611619
...

Sergei

vabadu commented 8 years ago

Yep, your results agree with mine.

Vanessa

From: Sergei Pond notifications@github.com Sent: Tuesday, February 16, 2016 2:58 PM To: veg/hyphy Cc: Vanessa Lamae Bauer DuMont Subject: Re: [hyphy] interpreting aBSREL results (#391)

Dear @vabaduhttps://github.com/vabadu,

Just to confirm, the equivalent analysis executed in HyPhy (https://github.com/veg/hyphy-tutorials/tree/master/selection#use-absrel-to-find-lineages-which-have-experienced-episodic-diversification) yields the following output:

... Summary of branches under episodic selection (7 were tested, of which 4 required optimizations) : Node4 p = 4.993840995837218e-06 WHITESOUTHERNRHINO p = 0.007700264980029792 DONKEY_TRIM25_1 p = 0.03754281322611619 ...

Reply to this email directly or view it on GitHubhttps://github.com/veg/hyphy/issues/391#issuecomment-184891449.

spond commented 8 years ago

Hi @vabadu,

OK, while we work on a fix for test.datamonkey.org, please run aBSREL locally, and use the web-app (http://veg.github.io/hyphy-vision/absrel/) to view the JSON file.

Sergei

JaneZXJ commented 8 years ago

Hi @spond, I met the same problem that "not tested" for each branch in the aBSREL results, and I'm not so clear about your resolution. [http://test.datamonkey.org/absrel/56c5806f9dd92ef47c709304] [http://test.datamonkey.org/absrel/56c5a9569dd92ef47c7093ad] And more questions about:

  1. Did it mean the results of PAML, BSREL and aBSREL should be in concordance? If not, as my BSREL results indeed detected episodic selected branches, but not pass LRT for rBS model in PAML.
  2. What's the difference between local HyPhy and web-server? Why did it work locally but not on web-server?
  3. I don't understand correlation of the results between GAB and BSR, as the dN/dS ratio of same lineage or branch is totally different between the two models. And their algorithms are different.
  4. So, after BSREL, because the branch results are in accordance with my expection, where is the site results? Do I need to perform another one, as covarion model of Guindon et al. 2004 cited in your BSREL paper? Or from the results of site-by-site selection? As there are same residues between HyPhy and PAML site models.

So appreciated for your time and any suggestion! Jane P.S. my data are hemoglobin genes of birds adapted to different elevations, if it's different from your familiar data.

spond commented 8 years ago

Dear @JaneZXJ,

Please see my responses to #394

Sergei

stevenweaver commented 8 years ago

This issue has been resolved and pushed to test.datamonkey.org.

Two adjustments were made :

1) The web version defaulted to (1):[Yes] Both alpha and beta vary along branch-site combinations in response to the prompt Allow branch-site variation in synonymous rates? The default has now been changed to (2):[No] [Default] Alpha varies from branch to branch, while omega varies among branch-site combinations.

2) test.datamonkey.org was using a version of aBSREL that was included in HyPhy tagged version 2.2.6. aBSREL has been updated to use the version in HyPhy commit cdea3c249f6b2a7dca7db9e024a7388ffe9c734f.

Best, Steven

JaneZXJ commented 8 years ago

Hi Steve, Is this the response to #389? about pdf output of BGM and GAB. If that, did you mean to run them at test.datamonkey.org?

     But the link seems unavailable, like 

        My brouser in Chrome.
        Thanks again!

Regards, Jane

Xiaojia Zhu Ph.D student Ornithological research group Key Laboratory of Zoological Systematics and Evolutionary Center 5 Institute of Zoology, CAS 1 Bei Chen West road Chao Yang District Beijing,100101 xiaojia0402@hotmail.com zhuxiaojia@ioz.ac.cn Office: 010-64807188 Phone: 86-13581827838

From: Steven Weaver Date: 2016-02-23 06:51 To: veg/hyphy CC: JaneZXJ Subject: Re: [hyphy] interpreting aBSREL results (#391) This issue has been resolved and pushed to test.datamonkey.org. Two adjustments were made : 1) The web version defaulted to (1):[Yes] Both alpha and beta vary along branch-site combinations in response to the prompt Allow branch-site variation in synonymous rates? The default has now been changed to (2):[No] [Default] Alpha varies from branch to branch, while omega varies among branch-site combinations. 2) test.datamonkey.org was using a version of aBSREL that was included in HyPhy tagged version 2.2.6. aBSREL has been updated to use the version in HyPhy commit cdea3c2. Best, Steven — Reply to this email directly or view it on GitHub.

DDrabeck commented 6 years ago

Hi! It looks like I am having a similar issue here. Results from aBSREL here: http://datamonkey.org/absrel/5afb3c9df26c344f314a2001

While there are several branches with w>1 it seems only one has reported a LRT, the remainder say "test not run". It looks like this string is older, so I thought it might be better to ask again in light of any updates.

Secondarily, I submitted several jobs yesterday afternoon and aBSREL is the only one I've gotten back. Wondering if this is because of the number of taxa (137 for alignment length of 702), or if something is wrong.

THANKS!

~Danielle

stevenweaver commented 6 years ago

Dear @DDrabeck,

Thank you for reporting this. I will investigate shortly.

Best, Steven

DDrabeck commented 6 years ago

Hi Steven,

I have another question about what seems to be a big issue... aBSREL does not have a prompt for uploading a tree file... it seems to only have an alignment file prompt that then generates a tree from the alignment.

Is there a fix for this? It seems clear in the description that this uses a tree, and generating a tree from sequences that have heavy selection is highly problematic... so hopefully there is a way to incorporate a species tree?

Analysis Description

aBSREL (Adaptive branch-site random effects likelihood) uses an adaptive random effects branch-site model framework to test whether each branch has evolved under positive selection, using a procedure which infers an optimal number of rate categories per branch.

On Wed, May 16, 2018 at 2:32 PM Steven Weaver notifications@github.com wrote:

Dear @DDrabeck https://github.com/DDrabeck,

Thank you for reporting this. I will investigate shortly.

Best, Steven

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/veg/hyphy/issues/391#issuecomment-389639007, or mute the thread https://github.com/notifications/unsubscribe-auth/AOcYPMVg-DIFoG0JfLbLUAiV3YTlr8Hgks5tzH7ngaJpZM4Hbixm .

-- Danielle H Drabeck M.Sc.

PhD CandidateDepartment of Ecology, Evolution, and BehaviorUniversity of Minnesota

Drabe004@umn.edu Danielle.Drabeck@gmail.com

stevenweaver commented 6 years ago

Dear @DDrabeck,

Datamonkey looks for the Newick tree inside of the multiple sequence alignment (FASTA) file. Please try appending it to the bottom of the FASTA file and upload.

Best, Steven

DDrabeck commented 6 years ago

Ah! Thank you!

On Thu, May 17, 2018 at 4:02 PM Steven Weaver notifications@github.com wrote:

Dear @DDrabeck https://github.com/DDrabeck,

Datamonkey looks for the Newick tree inside of the multiple sequence alignment (FASTA) file. Please try appending it to the bottom of the FASTA file and upload.

Best, Steven

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/veg/hyphy/issues/391#issuecomment-390010327, or mute the thread https://github.com/notifications/unsubscribe-auth/AOcYPE1QQHaJlTC9ZjyqVfe1ukxVfCOdks5tzeVUgaJpZM4Hbixm .

-- Danielle H Drabeck M.Sc.

PhD CandidateDepartment of Ecology, Evolution, and BehaviorUniversity of Minnesota

Drabe004@umn.edu Danielle.Drabeck@gmail.com

Bio-nic commented 5 years ago

Dear @stevenweaver, dear @spond, I have a problem interpreting the aBSREL results below (.json file). I want to use the "Baseline MG94xREV omega ratio" to get an idea of the mean dN/dS of each branch, but all branches with only one rate distribution (100%) show a value of 10,000,000,000 which is not what i expect. How should I interpret these values and is there an algorithm in HYPHY calculating the mean dN/dS for each branch of an alignment?

"BauWigWesBlo":{
       "Nucleotide GTR":0.1041013266186939,
       "Baseline MG94xREV":0.03538238627233805,
       "Baseline MG94xREV omega ratio":10000000000,
       "Rate classes":1,
       "Full adaptive model":1.692413846676369,
       "Rate Distributions":[
        [0.007957805752180375, 1]
        ],
       "LRT":0,
       "Uncorrected P-value":1,
       "Corrected P-value":1
      },
     "Origin2":{
       "Nucleotide GTR":0.08247267261885365,
       "Baseline MG94xREV":1068.746430561004,
       "Baseline MG94xREV omega ratio":1.304728023232953e-05,
       "Rate classes":2,
       "Full adaptive model":448.2182936838265,
       "Rate Distributions":[
        [0.001152201003741379, 0.9420913319677477],
        [596.2688599266926, 0.05790866803225225]
        ],
       "LRT":-0.03190303538940498,
       "Uncorrected P-value":0.5,
       "Corrected P-value":1
      },
     "WesBlo":{
       "Nucleotide GTR":0.0003199043849954112,
       "Baseline MG94xREV":0.01415919605646151,
       "Baseline MG94xREV omega ratio":10000000000,
       "Rate classes":1,
       "Full adaptive model":3.788584127487181,
       "Rate Distributions":[
        [0.001361061761396844, 1]
        ],
       "LRT":0,
       "Uncorrected P-value":1,
       "Corrected P-value":1
stevenweaver commented 5 years ago

Dear @spondfanboy,

To be clear, in aBSREL models, the baseline omega is meant to serve as a starting value for the optimization for aBSREL, and should not be used for inference.

With that said, the number 10000000000 is an infinite bound in HyPhy. This would be possible if dS = 0.

Best, Steven

spond commented 5 years ago

Dear @spondfanboy,

I am flattered by your handle. Not sure I can handle the fame and the glory though.

For the MEAN dN/dS using aBSREL I would recommend you use the mean of the full model distribution, which is included in the Rate Distributions key of the JSON object. @stevenweaver is correct in that the other values are initial approximations, and do not reflect the best description of the data among the models considered.

Best, Sergei