SEQBOOT -- Bootstrap, Jackknife, or Permutation Resampling
of Molecular Sequence, Restriction Site,
Gene Frequency or Character Data

<DIV ALIGN=RIGHT>
version 3.6
</DIV>
<P>
<DIV ALIGN=CENTER>
<H1>SEQBOOT -- Bootstrap, Jackknife, or Permutation Resampling<BR>
of Molecular Sequence, Restriction Site,<BR>
Gene Frequency or Character Data</H1>
</DIV>
<P>
&#169; Copyright 1991-2002 by the University of Washington.
Written by Joseph Felsenstein.  Permission is granted to copy 
this document provided that no fee is charged for it and that this copyright 
notice is not removed. 
<P>
SEQBOOT is a general bootstrapping and data set translation tool.  It is intended to allow you to
generate multiple data sets that are resampled versions of the input data
set.  Since almost all programs in the package can analyze these multiple
data sets, this allows almost anything in this package to be bootstrapped,
jackknifed, or permuted.  SEQBOOT can handle molecular sequences,
binary characters, restriction sites, or gene frequencies.  It
can also convert data sets between Sequential and Interleaved
format, and into NEXUS and a new XML sequence alignment format.
<P>
To carry out a bootstrap (or jackknife, or permutation test) with some method
in the package, you may need to use three programs.  First, you need to run
SEQBOOT to take the original data set and produce a large number of
bootstrapped or jackknifed data
sets (somewhere between 100 and 1000 is usually adequate).
Then you need to find the phylogeny estimate for
each of these, using the particular method of interest.  For example, if
you were using DNAPARS you would first run SEQBOOT and make a file with 100
bootstrapped data sets.  Then you would give this file the proper name to
have it be the input file for DNAPARS.  Running DNAPARS with the M (Multiple
Data Sets) menu choice and informing it to expect 100 data sets, you
would generate a big output file as well as a treefile with the trees from
the 100 data sets.  This treefile could be renamed so that it would serve
as the input for CONSENSE.  When CONSENSE is run the majority rule consensus
tree will result, showing the outcome of the analysis.
<P>
This may sound tedious, but the run of CONSENSE is fast, and that of
SEQBOOT is fairly fast, so that it will not actually take any longer than
a run of a single bootstrap program with the same original data and the same
number of replicates.  This is not very hard and allows bootstrapping on many of
the methods in
this package.  The same steps are necessary with all of them.  Doing things
this way some of the intermediate files (the tree file from the DNAPARS
run, for example) can be used to summarize the results of the bootstrap in
other ways than the majority rule consensus method does.
<P>
If you are using the Distance Matrix programs, you will have to add one extra
step to this, calculating distance matrices from each of the replicate data
sets, using DNADIST or GENDIST.  So (for example) you would run SEQBOOT, then
run DNADIST using the output of SEQBOOT as its input, then run (say) NEIGHBOR
using the output of DNADIST as its input, and then run CONSENSE using the
tree file from NEIGHBOR as its input.
<P>
The resampling methods available are three:
<UL>
<LI><B>The bootstrap.</B>  Bootstrapping was invented by Bradley Efron in 1979,
and its use in phylogeny estimation was introduced by me (Felsenstein, 1985b;
see also Penny and Hendy, 1985).
It involves creating a new data set by sampling <I>N</I> characters randomly
with replacement, so that the resulting data set has the same size as the
original, but some characters have been left out and others are duplicated.
The random variation of the results from analyzing these bootstrapped
data sets can be shown statistically to be typical of the variation that
you would get from collecting new data sets.  The method assumes that the
characters evolve independently, an assumption that may not be realistic
for many kinds of data.
<P>
<LI><B>Block-bootstrapping.</B>  One pattern of departure from indeopendence
of character evolution is correlation of evolution in adjacent characters.
When this is thought to have occurred, we can correct for it by samopling,
not individual characters, but blocks of adjacent characters.  This is
called a block bootstrap and was introduced by K&uuml;nsch (1989).  If the
correlations are believed to extend over some number of characters, you
choose a block size, <I>B</I>, that is larger than this, and choose
<I>N/B</I> blocks of size <I>B</I>.  In its implementation here the
block bootstrap "wraps around" at the end of the characters (so that if a
block starts in the last&nbsp; <I>B-1</B> characters, it continues by wrapping
around to the first character after it reaches the last character).  Note also
that if you have a DNA sequence data set of an exon of a coding region, you
can ensure that equal numbers of first, second, and third coding positions
are sampled by using the block bootstrap with <I>B = 3</B>.
<P>
<LI><B>Delete-half-jackknifing</B>.  This alternative to the bootstrap involves
sampling a random half of the characters, and including them in the data
but dropping the others.  The resulting data sets are half the size of the
original, and no characters are duplicated.  The random variation from
doing this should be very similar to that obtained from the bootstrap.
The method is advocated by Wu (1986).  It was mentioned by me in my
bootstrapping paper (Felsenstein, 1985b), and has been available for many
years in this program as an option.  Jackknifing is advocated by
Farris et. al. (1996) but as deleting a fraction 1/e (1/2.71828).  This
retains too many characters and will lead to overconfidence in the
resulting groups.
<P>
<LI><B>Permuting species within characters.</B>  This method of resampling (well, OK,
it may not be best to call it resampling) was introduced by Archie (1989)
and Faith (1990; see also Faith and Cranston, 1991).  It involves permuting the
columns of the data matrix
separately.  This produces data matrices that have the same number and kinds
of characters but no taxonomic structure.  It is used for different purposes
than the bootstrap, as it tests not the variation around an estimated tree
but the hypothesis that there is no taxonomic structure in the data: if
a statistic such as number of steps is significantly smaller in the actual
data than it is in replicates that are permuted, then we can argue that there
is some taxonomic structure in the data (though perhaps it might be just a
pair of sibling species).
</UL>
<P>
The data input file is of standard form for molecular sequences (either in
interleaved or sequential form), restriction sites, gene frequencies, or
binary morphological characters.
<P>
When the program runs it first asks you for a random number seed.  This should
be an integer greater than zero (and probably less than 32767) and which is
of the form 4n+1, that is, it leaves a remainder of 1 when divided by 4.  This
can be judged by looking at the last two digits of the integer (for instance
7651 is not of form 4n+1 as 51, when divided by 4, leaves the remainder 3).
The random number seed is used to start the random number generator.
If the randum number seed is not odd, the program will request it again.
Any odd number can be used, but may result in a random number sequence that
repeats itself after less than the full one billion numbers.  Usually this
is not a problem.  As the random numbers appear to be unpredictable,
there is no such thing as a "good" seed -- the numbers produced from one
seed are indistinguishable from those produced by another, and it is
not true that the numbers produced from one seed (say 4533) are similar to
those produced from a nearby seed (say 4537).
<P>
Then the program shows you a menu to allow you to choose options.  The menu
looks like this:
<P>
<TABLE><TR><TD BGCOLOR=white>
<PRE>

Bootstrapping algorithm, version 3.6a3

Settings for this run:
  D      Sequence, Morph, Rest., Gene Freqs?  Molecular sequences
  J  Bootstrap, Jackknife, Permute, Rewrite?  Bootstrap
  B      Block size for block-bootstrapping?  1 (regular bootstrap)
  R                     How many replicates?  100
  W              Read weights of characters?  No
  C                Read categories of sites?  No
  F     Write out data sets or just weights?  Data sets
  I             Input sequences interleaved?  Yes
  0      Terminal type (IBM PC, ANSI, none)?  (none)
  1       Print out the data at start of run  No
  2     Print indications of progress of run  Yes

  Y to accept these or type the letter for one to change

</PRE>
</TD></TR></TABLE>
<P>
The user selects options by typing one of the letters in the left column,
and continues to do so until all options are correctly set.  Then the
program can be run by typing Y.
<P>
It is important to select the correct data type (the D selection).  Each
time D is typed the program will change data type, proceeding successively
through Molecular Sequences, Discrete Morphological Characters, Restriction
Sites, and Gene Frequencies.  Some of these will cause additional entries
to appear in the menu.  If Molecular Sequences or Restriction Sites settings
and chosen the I (Interleaved)
option appears in the menu (and as Molecular Sequences are also the default,
it therefore appears in the first menu).  It is the usual
I option discussed in the Molecular Sequences document file and in the main
documentation files for the package, and is on by default.
<P>
If the Restriction Sites option is chosen the menu option E appears, which
asks whether the input file contains a third number on the first line of
the file, for the number of restriction enzymes used to detect these sites.
This is necessary because data sets for RESTML need this third number, but
other programs do not, and SEQBOOT needs to know what to expect.
<P>
If the Gene Frequencies option is chosen an menu option A appears which allows
the user to specify that all alleles at each locus are in the input file.
The default setting is that one allele is absent at each locus.
<P>
The J option allows the user to select Bootstrapping, Delete-Half-Jackknifing,
or the Archie-Faith permutation of species within characters.  It changes
successively among these three each time J is typed.
<P>
The B option selects the Block Bootstrap.  When you select option B the program
will ask you to enter the block length.  When the block length is 1,
this means that we are doing regular bootstrapping rather than
block-bootstrapping.
<P>
The R option allows the user to set the number of replicate data sets.
This defaults to 100.  Most statisticians would be happiest with 1000 to
10,000 replicates in a bootstrap, but 100 gives a rough picture.  You
will have to decide this based on how long a running time you are willing to
tolerate.
<P>
The W (Weights) option allows weights to be read
from a file whose default name is "weights".  The weights
follow the format described in the main documentation file.
Weights can only be 0 or 1, and act to select
the characters (or sites) that will be used in the resampling, the others
being ignored and always omitted from the output data sets.
<B>Note:</B> At present, if you use W together with the F (just weights)
option, you write a file of weights, but with only weights for the
sites that had input weights of 1, the others being omitted.  Thus if
you had 100 characters, and gave 60 of them weights of 1, when you
produce the output weights these will only have 60 weights, not 100.
Thus they could only be used together with a data file that had been
edited to remove the sites that you gave 0 weights to.  This is
clumsy and we need to correct it.
<P>
The C (Categories) option can be used with molecular sequence programs to
allow assignment of sites or amino acid positions to user-defined rate
categories.  The assignment of rates to
sites is then made by reading a file whose default name is "categories".
It should contain a string of digits 1 through 9.  A new line or a blank
can occur after any character in this string.  Thus the categories file
might look like this:
<P>
<PRE>
122231111122411155
1155333333444
</PRE>
<P>
The only use of the Categories information in SEQBOOT is that they
are sampled along with the sites (or amino acid positions) and are
written out onto a file whose default name is "outcategories",
which has one set of categories information for each bootstrap
or jackknife replicate.
<P>
The F option is a particularly important one.  It is used whether to
produce multiple output files or multiple weights.  If your
data set is large, a file with (say) 1000 such data sets can be very
large and may use up too much space on your system.  If you choose
the F option, the program will instead produce a weights file with
multiple sets of weights.  The default name of this file is "outweights".
Except for some programs that cannot handle multiple sets of
weights,
the programs have an M (multiple data sets) option that asks the
user whether to use multiple data sets or multiple sets of weights.
If the latter is selected when running those programs, they
read one data set, but analyze it multiple times, each time reading a new
set of weights.  As both bootstrapping and jackknifing can be thought of
as reweighting the characters, this accomplishes the same thing (the
multiple weights option is not available for Archie/Faith permutation).
As the file with multiple sets of weights is much smaller than a file with
multiple data sets, this can be an attractive way to save file space.
When multiple sets of weights is chosen, they reflect the sampling as
well as any set of weights that was read in, so that you can use
SEQBOOT's W option as well.
<P>
The 0 (Terminal type) option is the usual one.
<P>
<H2>Input File</H2>
<P>
The data files read by SEQBOOT are the standard ones for the various kinds of
data.  For molecular sequences the sequences may be either interleaved or
sequential, and similarly for restriction sites.  Restriction sites data
may either have or not have the third argument, the number of restriction
enzymes used.  Discrete morphological
characters are always assumed to be in sequential format.  Gene frequencies
data start with the number of species and the number of loci, and then
follow that by a line with the number of alleles at each locus.  The data for
each locus may either have one entry for each allele, or omit one allele at
each locus.  The details of the formats are given in the main documentation
file, and in the documentation files for the groups of programs.
<P>
The only option that can be present in the
input file is F (Factors), the latter only in the case of
binary (0,1) characters.  The Factors
option allows us to specify that groups of binary characters represent
one multistate character.  When sampling is done they will be sampled or
omitted together, and when permutations of species are done they will all
have the same permutation, as would happen if they really were just one
column in the data matrix.  For futher description of the F (Factors) option
see the Discrete Characters Programs documentation file.
<P>
<H2>Output</H2>
<P>
The output file will contain the data sets generated by the resampling
process.  Note that, when Gene Frequencies data is used or when
Discrete Morphological characters with the Factors option are used,
the number of characters in each data set may vary.  It may also vary
if there are an odd number of characters or sites and the Delete-Half-Jackknife
resampling method is used, for then there will be a 50% chance of choosing
(n+1)/2 characters and a 50% chance of choosing (n-1)/2 characters.
<P>
The order of species in the data sets in the output file will vary
randomly.  This is a precaution to help the programs that analyze these data
avoid any result which is sensitive to
the input order of species from showing up repeatedly
and thus appearing to have evidence in its favor.
<P>
The numerical options 1 and 2 in the menu also affect the output file.
If 1 is chosen (it is off by default) the program will print the original
input data set on the output file before the resampled data sets.  I cannot
actually see why anyone would want to do this.  Option 2 toggles the
feature (on by default) that prints out up to 20 times during the resampling
process a notification that the program has completed a certain number of
data sets.  Thus if 100 resampled data sets are being produced, every 5
data sets a line is printed saying which data set has just been completed.
This option should be turned off if the program is running in background and
silence is desirable.  At the end of execution the program will always (whatever
the setting of option 2) print
a couple of lines saying that output has been written to the output file.
<P>
<H2>Size and Speed</H2>
<P>
The program runs moderately quickly, though more slowly when the Permutation
resampling method is used than with the others.
<P>
<H2>Future</H2>
<P>
I hope in the future to include code to pass on the Ancestors
option from the input file (for use in programs MIX and DOLLOP)
to the output file, a serious
omission in the current version.
<P>
<HR>
<P>
<H3>TEST DATA SET</H3>
<P>
<TABLE><TR><TD BGCOLOR=white>
<PRE>
    5    6
Alpha     AACAAC
Beta      AACCCC
Gamma     ACCAAC
Delta     CCACCA
Epsilon   CCAAAC
</PRE>
</TD></TR></TABLE>
<P>
<HR>
<P>
<H3>CONTENTS OF OUTPUT FILE</H3>
<P>
(If Replicates are set to 10 and seed to 4333) 
<P>
<TABLE><TR><TD BGCOLOR=white>
<PRE>
    5     6
Alpha     ACAAAC
Beta      ACCCCC
Gamma     ACAAAC
Delta     CACCCA
Epsilon   CAAAAC
    5     6
Alpha     AAAACC
Beta      AACCCC
Gamma     CCAACC
Delta     CCCCAA
Epsilon   CCAACC
    5     6
Alpha     ACAAAC
Beta      ACCCCC
Gamma     CCAAAC
Delta     CACCCA
Epsilon   CAAAAC
    5     6
Alpha     ACCAAA
Beta      ACCCCC
Gamma     ACCAAA
Delta     CAACCC
Epsilon   CAAAAA
    5     6
Alpha     ACAAAC
Beta      ACCCCC
Gamma     ACAAAC
Delta     CACCCA
Epsilon   CAAAAC
    5     6
Alpha     AAAACA
Beta      AAAACC
Gamma     AAACCA
Delta     CCCCAC
Epsilon   CCCCAA
    5     6
Alpha     AAACCC
Beta      CCCCCC
Gamma     AAACCC
Delta     CCCAAA
Epsilon   AAACCC
    5     6
Alpha     AAAACC
Beta      AACCCC
Gamma     AAAACC
Delta     CCCCAA
Epsilon   CCAACC
    5     6
Alpha     AAAAAC
Beta      AACCCC
Gamma     CCAAAC
Delta     CCCCCA
Epsilon   CCAAAC
    5     6
Alpha     AACCAC
Beta      AACCCC
Gamma     AACCAC
Delta     CCAACA
Epsilon   CCAAAC
</PRE>
</TD></TR></TABLE>
<P>