Difference between revisions of "Migrate tutorial 2015"

From MolEvol
(Migration/Divergence matrix modifications)
Line 214: Line 214:
  
 
The connection matrix is square so we can label it like it is shown in the first table below.
 
The connection matrix is square so we can label it like it is shown in the first table below.
{| class="wikitable" style="float; margin-left: 10px;border:1px solid #BBB" cellpadding=5
+
{| class="wikitable" style="float:right; margin-left: 10px;border:1px solid #BBB" cellpadding=5
 
|+ Table 1: Model for uni-directional migration
 
|+ Table 1: Model for uni-directional migration
 
|- bgcolor="#fafeff"
 
|- bgcolor="#fafeff"
Line 231: Line 231:
 
|Brissago || *  || *  
 
|Brissago || *  || *  
 
|}
 
|}
 
 
  
 
* MIGRATE asks now that you input each row, this can be done by either specifying '''* 0''' (see second table) and then return and then entering the next line '''* *''' return (second row in second table), or you can enter the whole matrix as '''* 0 * *'''.
 
* MIGRATE asks now that you input each row, this can be done by either specifying '''* 0''' (see second table) and then return and then entering the next line '''* *''' return (second row in second table), or you can enter the whole matrix as '''* 0 * *'''.

Revision as of 19:51, 22 July 2015

Tutorial Overview

How to get MIGRATE

Migrate2logo.png

Locally, the most recent copy of the software on the server is in the /class/molevol-software/migrate directory, you can either use the program that is already installed on the class server or use these files in the local_distribution directory to install MIGRATE on your own computer. if this fails, download MIGRATE from which is the Migrate download-website.

Comparison of gene flow models using Bayes Factors with MIGRATE

thump

Most are familiar with the concept of likelihood ratio tests, or Akaike’s Information criterion for model comparison. This tutorial describes how to compare population models using Bayes Factors. These allow comparing nested and un-nested models, without assuming Normality, or large samples. Bayes factors are ratios of marginal likelihoods. In contrast to the maximum likelihood, the marginal likelihood is the integral of the likelihood function over the complete parameter range. MIGRATE can calculate such marginal likelihoods for a particular migration model (Beerli and Palczewski 2010).

Executive Summary

This tutorial steps through all necessary program runs to calculate Bayes factors for comparing different gene flow models. We need to do the following:

  • Decide on the models that are interesting for a comparison. The method does not work well for a fishing expedition where one would try to evaluate all possible models; this is only possible only for population models with very few number of populations. It will be possible to enumerate all models for two or three populations but more will be very daunting.
  • Run each model through MIGRATE. Use the same prior settings for each of them because the prior distribution has some influence on the Bayes factors. Use the heating menu to allow for at least four chains. The menu supplies a shortcut to specify the temperatures, it is #. It generates temperatures that are spaced in a particular way: they are spaced so that the inverse of the temperature are regularly spaced on the interval 0 to 1. For example, the 4 different chains have temperatures 1.0, 1.5, 3.0, 100,000.0, this results in the spacing 1.0, 0.666, 0.333, and 0.0.
  • Compare the marginal likelihood of the different runs and calculate the Bayes factor and calculate the probability for each model.


Extensive Details

The following pages detail all steps using a small example. We use a simulated dataset that was generated using parameters that force a direction of migration from the population Ascona (A) to the population Brisago (B). The Brissago population is larger than the Ascona population and no individual from Brissago ever goes to Ascona, but Brissago receives about 1 migrant every four generation from Ascona. The dataset name is twoswisstowns (if you are not at the workshop then download here) We will evaluate a total of 7 population models.

  1. a full migration model with two population sizes and two migration rates (from A to B and from B to A);
  2. a model with two population sizes and one migration rate to Brissago;
  3. a model with two population sizes and one migration rate to Ascona;
  4. a model where Ascona and Brissago are part of the same panmictic population.
  5. a model where Ascona is an ancient city and Brissago was built a new from people who left Ascona; but Brissago is still attractive and people migrate to there ever since.
  6. a model where Brissago is an ancient city and Ascona was built a new from people who left Brissago; but Ascona is still attractive and people migrate to there ever since.
  7. a model where Ascona is an ancient city and Brissago was built a new from people who left Ascona and they became hostile towards each other and nobody moves between the cities.

. We know the truth therefore we have some prejudice about the ranking of the models, model 5 should be best, models with the same migration pattern as the truth should work better than those with alternative migration patterns, or those with many parameters.

First we need to figure out how to run the dataset efficiently in MIGRATE. For that we

pick a parameter-rich model 1 and experiment with run conditions until we are satisfied that the run converges and delivers posterior distributions that look acceptable. Here are now the detailed instructions how to rank population genetics models for a particular dataset.


Familiarize with MIGRATE [Tutorial Start]

  • Make a new directory and download or copy the datafile
#if you are AT THE WORKSHOP use this
rsync -avz /class/molevol-software/migrate/migrate_tutorial .
cd migrate_tutorial
#if you are NOT at the workshop use this
mkdir migrate_lab
cd migrate_lab
wget http://popgen.sc.fsu.edu/tutorials/BF_migrate_tutorial2/twoswisstowns
  • Start the program: the regular distribution comes in two flavors the single cpu processor version called migrate-n and the parallel processing version that runs on cluster or computers with multiple cores is called migrate-n-mpi. We use a new version of MIGRATE: version 4.2.2. On the cluster you should be able to call migrate-n. (In this text I will call the program from now on simply MIGRATE).

We will run the exercise on the server. On the server type

migrate-n  

On your laptop you may need to use "./migrate-n", if the program is in the same directory. The main menu will appear, looking like this

[pbeerli@class-02 migrate_lab]$ migrate-n

 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 +                                                                +
 +   POPULATION SIZE, MIGRATION, DIVERGENCE, ASSIGNMENT, HISTORY  +
 +   Bayesian inference using the structured coalescent           +
 +                                                                +
 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
  PDF output enabled [Letter-size]
  Version 4.2.2a   [July-18-2015]
  Program started at   Tue Jul 21 15:03:15 2015


  =============================================================
  MAIN MENU
  =============================================================

  D       Data type currently set to: DNA sequence model            
  I       Input/Output formats and Event reporting
  P       Parameters  [start, migration model]
  S       Search strategy
  W       Write a parmfile
  Q       Quit the program


  To change the settings type the letter for the menu to change
  Start the program with typing Yes or Y
===> 

  • Go to the Input/Output formats menu (press I and hit Enter), in the INPUT section change the Datafile name to twoswisstowns, Return to the main menu by typing Y.
  • In the Search strategy menu: Change the Number of recorded steps in chain to 1000, and also change the Burn-in for each chain: to 1000. Do not worry about priors or other runtime options for the moment. Return to the main menu.
  • Save the changes by using the menu item Write a parmfile. This will write a file named parmfile.
  • Now run the program (pressing Y will start the run if you are in the main menu). For this dataset the runtime will be very short. On a modern computer this will take under a minute. If this takes more than 3 minute, something is not set up correctly! On the server this takes about 29.6 seconds (on my macOS 10.10.4: 8.8 seconds).
  • The program writes considerable information during the run to the screen, that gives some information about the run. Most interesting are the acceptance ratio for the genealogy and the autocorrelations of the parameter and the genealogy. If the autocorrelation is high and the effective sample size is low (<500) then a longer run may be needed. If the priors boundaries are to tight, then you will see that the values reported are either very close or exactly at the upper prior boundary, in these cases you need to extend the prior range. See prior problems in the output, but for this dataset we will have no such problems.
  • Look at the outfile.pdf, you will need to transfer the pdf file to your computer and use preview or acrobat or another PDF viewer. In the outputfile look at the figures labeled Bayesian analysis: Posterior distribution, you see histograms similar to the ones in Figure 1. We expect single peaks where the shading of the histogram shows one dark block in the center (50% credibility set), two light gray bars indicating the extent of the 95% credibility set, and two lighter gray bars indication the 99% credibility set. We expect a histrogram that looks smooth that usually has a single peak, commonly similar to an exponential or normal distribution with a heavy right tail.
Figure 1: Example if a short run showing an unsatifactory posterior histogram; red lines are the prior distributions used
  • In your investigation of Figure 1 you recognize that the histogram has a few kinks because our run was too short. Now restart MIGRATE and set in the strategy menu the setting for change the number of recorded steps in chain from 1,000 to 10,000. This will lengthen the run by a factor of 10.



Because we want to use the thermodynamic integration method for our marginal likelihood calculations, we need to turn on heating. Start MIGRATE, use the Strategy menu and turn on heating. MIGRATE will tell what to do next, you will need to enter 4 chains sampling at every tenth (10) interval using the temperature scheme that is suggested with the character #. Save the parmfile, and run. Here the sequence of entries that you will need to enter:

  ......
  13   Heating:                                                     NO
  16   Run analysis without data:                                   NO

  Are the settings correct?
  (Type Y to go back to the main menu or the number for a menu to change)

===>13

  Heating scheme? < NO | YES | STATIC | BOUNDED_ADAPTIVE >

===> Static

Enter the number of different "heated" chains.
Minimum is 4

===> 4

Enter the interval between swapping trees
Enter 0 (zero) for NO swapping
[Current interval is 1]

===> 1

 Enter 4 "temperatures"
[The coldest temperature, which is the first, has to be 1
 For example: 1.0 1.5 3.0 1000000.0]
OR give a range of values  [linear increase:      1 - 10]
                           [exponential increase: 1 @ 10]
or, most lazily, let me suggest a range [simply type a #]
@@@@@ For model comparison, the range of temperatures @@@@@
@@@@@ MUST include a very hot chain (>100000.0)       @@@@@

===> #

Chain         Temperature
-------------------------
   4        1000000.00000
   3              3.00000
   2              1.50000
   1              1.00000

Is this correct? [YES or NO]

===> yes

Don't forget to write the parmfile to save the settings. It should give a better posterior distribution histogram and will add a full table of (natural) log marginal likelihoods is shown towards the end of the outfile.pdf.

Run and compare the results with the histogram from before. Adding heating will take about 4x longer than before. You will recognize that the longer run has somewhat smoother histograms, but you may still see double peaks. With your own data you may want to do another round of refinements, but eventually, by comparing the medians and modes of the parameters in the table and the shape of the histograms you should see a good agreement on similar values, if the modes of the different runs are not within the 50% credibility intervals you certainly need to run longer. My run on the cluster took about 1517 seconds, on my mac it takes 275 seconds, but we could run this in parallel, by calculating the parameters for the loci in parallel, for example instead of

migrate-n parmfile
<pre>
we could execute this (these instructions will work on the MBL cluster and may most likely not work on your computer or other clusters).
<pre style="font-size: 1.25em">
module load openmpi
mpirun -np 6 migrate-n-mpi parmfile

Run this way, it will set up a computer with 6 virtual nodes (one master and 5 workers); the workers will do most of the work and at the end report back to the master who then assembles and prints the results. On the cluster, this takes 339 seconds (on my mac 149 seconds).

Finally, let's work on the different models

We now start to work on the different models. Remember, we want to run 7 models, and this will take some time, I suggest that you work with your neighbor and also use the parallel version.To make sure that we have examples of all different models in class start with the model that is equivalent to the class computer name you use, for example I am on class07 start with model 7. If you are on class08, class09, class10 pick model 5, 6, or 7, respectively. Once you finished your designates models work on the other ones, the more models the better.

Create directories for all model

I have created a small script that assisst with the creation of all model directories, execute at the prompt:

[pbeerli@class-02 migrate_lab]$ migrate-n
create_migmodels

It will create the directories and will also copy the parmfile into all these directories because we will need to change each parmfile to accommodate the population model.

Model 1: xxxx

Xxxx.png

This model allows migration between A and B, the populations are assumed to exist since a very long time. The xxxx in the title is a shortcut for this model it means that we estimate migration rates and population sizes. I use a particular notion to define the model, assume there is a migration matrix (or call it population interaction matrix), on the diagonal we have instructions for the population size, for example, estimate (x or *) or constant (c), there are several other possibilities which you could look up in the manual. The off-diagonal are the immigration rates, a * or 'x' means that this parameter is estimated. This particular model is the default and, in principle doe not need any further input and you can simply run MIGRATE, to report follow now the instructions in section Reporting values.

Model 2: x0xx

X0xx.png

This model allows migration from A to B. The populations are assumed to exist since a very long time. The x0xx in the title is a shortcut for this model it means that we estimate an immigration rate into the second population and population sizes. I use a particular notion to define the model, read the [#offdiag | instructions about the migration/divergence matrix] completely before you start typing.


Exit the parameter menu, write the parmfile, run MIGRATE to report follow now the instructions in section Reporting values.


Model 3: xx0x

Model 4: x

X.png
Relabel menu.png

Start MIGRATE and choose the menu Parameter settings. Choose the entry about sampling locations. We want to use the data as if we would have sampled a single population, therefore we need to claim that the two locations Ascona and Brissago belong to the same panmictic population. MIGRATE's default is to assume that every location is a individual population. The dialog (figure on the right gives an example of this interaction with the menu) will ask first how many locations are in the dataset (for our example we have 2). After that, you will need to assign the locations to a population. For this model we need to assign each location to the same population, so location 1 (Ascona) belongs to population 1 and location 2 (Brissago) also belongs to population 1. You need to enter 1 1 (one space one). With multiple populations more complicated settings are possible. Save the parmfile and then run MIGRATE. Follow now the instructions in section Reporting values.

Model 5: x0Bx

Model 6: xB0x

Model 7: x0dx

Report marginal likelihoods of Model 1 (4 Parameter)

[remember work with your neighbor, the runs will take a while and the best strategy to work together entering the options on one computer, starting the analysis, then update the options on the other computer and work on the next step in the tutorial]

Migration/Divergence matrix modifications

Start MIGRATE, choose the Parameter menu. Choose the entry labeled Model is set to. MIGRATE will now show a dizzying list of options, don't panic, we will only use a few of them. MIGRATE will ask you how many populations are used: enter 2. For a 2-population model we can have 4 parameters. For example, two population sizes and two migration rates. A * or x means that that particular parameter will be unrestrictedly estimated, a zero (0) means that that particular parameter will not be estimated (is not used). Our goal is to set one of the migration/divergence parameters to 0. MIGRATE needs to know how to treat all connections between the populations.

The connection matrix is square so we can label it like it is shown in the first table below.

Table 1: Model for uni-directional migration
To\From Ascona Brissago
Ascona Θ MB → A
Brissago MA → B Θ
To\From Ascona Brissago
Ascona * 0
Brissago * *
  • MIGRATE asks now that you input each row, this can be done by either specifying * 0 (see second table) and then return and then entering the next line * * return (second row in second table), or you can enter the whole matrix as * 0 * *.
Table2.png






  • Let's assume that our runs are all satisfactory. We turn now to the best estimation of the marginal likelihood to compare models.
Marginaltable.png
  • Come to the front and write down the log marginal likelihood into the spreadsheet (look at the example figure labeled Log-Probability of the data given model (marginal likelihood)) You will need the numbers from the row labeled All, in the table there are three columns, report the values for the Bezier approximation. (This was our first model, we will compare the different models at the end of this exercise: my log marginal likelihood values for the Bezier approximated score are -4803.07, respectively (see example figure on the right)
  • To save what we have done so far, copy the parmfile to parmfile.4param, copy outfile to outfile.4param, and copy outfile.pdf to outfile.4param.pdf. MIGRATE allows to specify filenames in the menu but copying in the terminal is fast and also allows us to use the parmfile as a template for the other models, so that we do not need to change the run-length and heating parameters again.
cp parmfile parmfile.4param
cp outfile.pdf outfile.4param.pdf
cp outfile outfile.4param

Report marginal likelihoods of Model 4!! (1 Parameter)

  • We start now to work on those other models. We pick the easiest first: model 4.
Relabel menu.png
  • Start MIGRATE and choose the menu Parameter settings. Choose the entry about sampling locations. We want to use the data as if we would have sampled a single population, therefore we need to claim that the two locations Aadorf and Bern belong to the same panmictic population. MIGRATE's default is to assume that every location is a individual population. The dialog (figure on the right gives an example of this interaction with the menu) will ask first how many locations are in the dataset (for our example we have 2). After that you will need to assign the locations to a population. For this model we need to assign each location to the same population. You need to enter 1 1 (one space one). With multiple populations more complicated settings are possible. Run MIGRATE, check the histogram, if it looks OK, come to the front and write down the log marginal likelihoods (again the row labeled All, Bezier and Harmonic score) into the spreadsheet under model 4. My run took 300 seconds and delivered the log marginal likelihood -4819.48.
  • Copy the parmfile to parmfile.1param, copy outfile to outfile.1param, and copy outfile.pdf to outfile.1param.pdf
cp parmfile parmfile.1param
cp outfile.pdf outfile.1param.pdf
cp outfile outfile.1param

Report marginal likelihoods of Model 2 (3 Parameter)

Figure3.png
  • Now you need to consider the models with unidirectional migration.
cp parmfile.4param parmfile
  • Start MIGRATE, choose the parameter menu. Choose the entry labeled Model is set to. MIGRATE will now show a dizzying list of options, don't panic, we will only use few of them. MIGRATE will ask you how many populations are used: enter 2. For a 2-population model we can have 4 parameters. Two population sizes and two migration rates. Before you enter values, please read this whole paragraph. A * or x means that that particular parameter will be unrestrictedly estimated, a zero (0) means that that particular parameter will not be estimated (is not used). Our goal is to set one of the migration/divergence parameters to zero. We start with model 2 (Figure 3). MIGRATE needs to know how to treat all connections between the populations. How migration rates or divergences are specified. We also need to give instructions how the program will treat the population sizes. Because we want to estimate both population sizes and one migration rate, we will use the * and a zero for the unused migration rate. The connection matrix is square so we can label it like it is shown in the first table below.
Table1.png
  • MIGRATE asks now that you input each row, this can be done by either specifying * 0 (see second table) and then return and then entering the next line * * return (second row in second table), or you can enter the whole matrix as * 0 * *.
Table2.png

Exit the parameter menu, write the parmfile, run MIGRATE, check the histogram, report the log marginal likelihoods. My run took 420 seconds, and delivered this log marginal likelihood: -4798.23. Cautionary note: if you use this tutorial for your own work, please recognize that a standard run in migrate can only use migration model that allow to draw a complete genealogy, so for example a model * 0 0 *, that has no migration among the population, does not work out of the box. As a rule of thumb each population must be connected to at least one other population.

  • Copy the parmfile to parmfile.3aparam, copy outfile to outfile.3aparam, and copy outfile.pdf to outfile.3aparam.pdf
cp parmfile parmfile.3aparam
cp outfile.pdf outfile.3aparam.pdf
cp outfile outfile.3aparam

Report marginal likelihoods of Model 3 (3 Parameter)

  • Run model 3 using the same procedure as for model 2. The string for the migration connection matrix is * * 0 *. Write parmfile, run, report. My run took 151 seconds and the log marginal likelihood was -4802.26 .
  • Copy the parmfile to parmfile.3bparam, copy out file to outfile.3bparam, and copy outfile.pdf to outfile.3bparam.pdf
  • Once about more than half of the class has reached this point we will talk about the marginal likelihoods found.
cp parmfile parmfile.3bparam
cp outfile.pdf outfile.3bparam.pdf
cp outfile outfile.3bparam

Report marginal likelihoods of Model 5 (3 Parameter)

  • Run model 5 using the same procedure as for model 3. The string for the connection matrix is * 0 d *.

In contrast to the models before this model will introduce a divergence event, the d marks the populaiton that split of from the ancestor, in our example this means Bern split off from Aadorf. Write parmfile, run, report. Report the log marginal likelihood; I got -4795.23.

  • Copy the parmfile to parmfile.3cparam, copy out file to outfile.3cparam, and copy outfile.pdf to outfile.3cparam.pdf
  • Once about more than half of the class has reached this point we will talk about the marginal likelihoods found.
cp parmfile parmfile.3cparam
cp outfile.pdf outfile.3cparam.pdf
cp outfile outfile.3cparam


Compare models

  • How to calculate Bayes factors? In the Table 3 I summarized all log marginal likelihoods, ln(mL), the Bayes factors are often calculated in very different ways. Here, I report the natural log Bayes factors where
Formula1.png
  • Using the guidelines of Kass and Raftery (1995), values smaller than -2 suggest preference for 'model 2', values larger than 2 suggest preference for 'model 1'. We can use the log marginal likelihoods or the BF to order the models (see column Choice in the Table 3).
  • We also can calculate the model probability. It is calculated by dividing each marginal likelihood by the sum of the marginal likelihoods of all used models:
Formula2.png
  • Note that for the above formula uses the marginal likelihoods, *not* the *log* marginal likelihoods (which is what the program reports). The calculation of model probabilities from the reported log likelihoods is easy with computer programs that have variable precision (for example Maple or Mathematica). Calculations on a desk calculator often fail, for example the likelihood of model 1 is a remarkable small number because the likelihood is exp(-4803.07)= 1.130323625060 x 10-2086, my emulated HP sci 15C calculator delivers 0.0000. But you can calculate the above quantities using this recipe: (1) find the largest log likelihood (-4795.23), (2) subtract that number form each log likelihood in the list (result: -2.27, 0.0, 2.5, -26.67), (3) exponentiate each element in the new list (result: 0.1033, 1.0, 0.0821, 2.6144 x 10-12) , (4) sum all elements in the list up (0.1033+1.0+0.0821+2.6144 x 10-12), this is the denominator (1.1854). (5) now divide each element in the list by that sum and the numbers will look like the one in table 3 last column.


Table 3: Showing log marginal likelihoods for all models tested and model prorabilities

Model Bezier lmL Choice Model probability
1: full (****) -4803.07 3 0.0077
2: true (*0**) -4798.23 1 0.9750
3: wrong (**0*) -4802.26 2 0.0173
4: panmictic (*) -4819.48 4 0.0000
5: divergence (*0d*) -4909.51 5 0.0000

Looking at the model probabilities we can see that the “true” model has considerably higher support than the full model or the model that suggests a wrong direction of gene flow.

Summary of results

The best model (the one with the highest marginal likelihood) is model 2 (custom-migration={*0**}) if we use the the thermodynamic approximation of marginal likelihood. MIGRATE also reports the harmonic mean, but I suggest to ingore it and use thermodynamic integration (as we did) although it will be more costly to run. I was shown several times (Beerli and Palczewski 2010, Xie et al. 2011 ) now that the harmonic mean estimator is not a good estimator and may be misleading and prefer the more complex model (like in our example). The picture below is a sample from the class tutorial done on August 1st 2011.

Migrate exercise.png


a second look at model comparison

there is a second dataset in the migrate tutorial, twoswisstownsdiv, here the sequence data was simulated using the divergence model, it is model 5 from the earlier excercise. Try tun compare the true divergence model (model 5, but change the infile option to the twoswisstownsdiv) and compare it to a run that uses model2.

cp parmfile.3cparam parmfile
# change the datafile to twoswisstownsdiv
#run migrate 
cp outfile.pdf outfileB1.pdf
cp outfile outfileB1
cp parmfile.3aparam parmfile
# change the datafile to twoswisstownsdiv 
#run migrate 
cp outfile.pdf outfileB2.pdf
cp outfile outfileB2
#compare the two models

I added a little python script that calculates the model probabilities from te text outfiles, you can run it using:

module load bioware
grep " All   " outfileB* | sort -n -k 4,4 | migbf.py

If you want to see the results of the first model exercise do:

grep " All   " outfile.* | sort -n -k 4,4 | migbf.py

References

  • Beerli, P. and M. Palczewski. 2010. Unified framework to evaluate panmixia and migration direction among multiple sampling locations. Genetics 185: 313–326.
  • Kass, R. E. and A. E. Raftery. Bayes factors. 1995. Journal of the American Statistical Association 90(430): 773– 795.
  • Xie, W., P. O. Lewis, Y. Fan, L. Kuo, and M.-H. Chen. 2011 Improving marginal likelihood estimation for Bayesian phylogenetic model selection, Systematic Biology, 60: 150–160.








Model Log(mL) LBF Model-probability


1:xxxx: -9662.42 -23.73 0.0000 2:xBxx: -9661.98 -23.29 0.0000 3:xxBx: -9661.52 -22.83 0.0000 4:xd0x: -9656.51 -17.82 0.0000 5:xB0x: -9649.33 -10.64 0.0000 6:xx0x: -9648.93 -10.24 0.0000 7:x0dx: -9641.77 -3.08 0.0402 8:x0xx: -9641.01 -2.32 0.0859 9:x0Bx: -9638.69 0.00 0.8739

Model: