TopDom() no longer produces a warning object 'idxs' not found
when called on data where no region can be processed, e.g. when
every non-gap run of bins is shorter than the minimum region size.
TopDom() with statFilter = TRUE (default) on extremely sparse
data used to fail deep inside stats::wilcox.test() with an
obscure not enough (finite) 'x' observations error. It now fails
early with an informative message explaining that the data is too
sparse for the statistical-filtering step and suggesting remedies
such as statFilter = FALSE or a larger bin size.
TopDom() with statFilter = TRUE (default) on a very large
contact matrix (more than sqrt(.Machine$integer.max), i.e. 46340,
bins) used to fail with an obscure 'to' must be a finite number
error from seq(), preceded by an NAs produced by integer
overflow warning, because the internal n_bins * n_bins index
arithmetic overflowed R's integer range. It now fails early with
an informative message suggesting statFilter = FALSE or a larger
bin size.
countsPerRegion(), subsetByRegion(), readHiC(),
ggCountHeatmap(), ggDomain(), ggDomainLabel(), legacy(), the
overlapScores() methods, several TopDom() options (statFilter,
outFile, debug, file-path input), and internal utility functions.Fix equation format issue and update one URL.
Drop stale and redirecting URLs (dead lab-software and USC library pages, PubMed/PMC links) in the help pages and README, keeping DOIs for the cited publications.
Package now includes mouse Chr19 data from the TopDom study. They
can be found in the system.file("exdata", package = "TopDom")
folder.
The orignal TopDom scripts TopDom_v0.0.1.R and TopDom_v0.0.2.R
are now distributed part of the package as-is. They can be found
in the system.file("original-scripts", package = "TopDom")
folder.
readHiC() gained arguments ... which is passed as-is to
read.table().
ggCountHeatmap() for TopDomData could produce a warning on a
partial argument name.overlapScores() now has
column chromosome as the first position. The data.frame:s are of
kind tibble.as_tibble() for TopDomOverlapScores.Add further documentation on the window.size parameter.
Add reference to Hanjun Shin's PhD thesis.
Improved help on overlapScores() and TopDom().
Provide a reference for the default value for window.size of
TopDom().
The TopDom object returned by TopDom() now has an attribute
parameters which records the value of arguments window.size and
statFilter.
Made TopDom() faster and more memory efficient by lower the
number of replicated computations.
TopDom() asserting that the
intermediate and final results are of proper length and does not
contain missing values.Convert.Bin.To.Domain.TMP() used by TopDom() could
produce Error in[<-.data.frame(tmp, , "to.coord", value =
c(NA, 2500, 247500 : replacement has 3 rows, data has 1, because
it assumed at least one domain was identified.overlapScores() to be in singular
form, e.g. best_score instead of best_scores.overlapScores() returns also the lengths of the reference
domains.overlapScores() and renamed the second argument to reference.
This was done in order to make it clear which set of topological
domains the overlap scores are calculated relative to.Add countsPerRegion() for calculating the total contact-frequency
counts per region specified, e.g. per domain.
Add print(), dim(), [(), and subsetByRegion() for TopDom
objects where the number of rows in the dimension reflect the
number of TopDom domains.
The legacy TopDom() functions, available via legacy(), also
accept TopDomData objects as returned by readHiC(). This is
supported mostly to make it possible to efficiently compare the
different implementations.
Added [() for TopDomData objects, e.g. tdd[1:100].
Added subsetByRegion() for TopDomData objects.
Added ggCountHeatmap(), ggDomain(), and ggDomainLabel() for
TopDomData objects.
legacy() for access to the original TopDom v0.0.1
and TopDom v0.0.2 implementations,
e.g. TopDom::legacy("0.0.1")$TopDom().TopDom() itself were the ones from
TopDom v0.0.2.Add overlapScores().
Add image() for TopDomData.
List returned by TopDom() gained class TopDom.
Added logical option TopDom.debug, which controls whether
functions produce debug output or not. The default is FALSE.
help("TopDom") with details from
the TopDom Manual (an online PDF) provided by Shin et al.Add print() method for TopDomData object.
Reference the TopDom paper (Shin et al., 2016) in the help and the README.
Turned the original TopDom R script into a package.
All progress messages are outputted done to standard error.
Add readHiC().
TopDom() can now read, via readHiC(), a pure count matrix file
without bin information. To read such files, specify what
chromosome is being read (argument chr) and the bin size of the
count matrix (argument binSize).
If the matrix file is not of a known format, then TopDom()
produces an informative error. Previously it gave a message on
stdout and returned 0.
TopDom v0.0.2 script from http://zhoulab.usc.edu/TopDom/ with the below entries from the official release note:
Gap Identification module is changed.
Minor bug related to Change Points identification in very small regions is fixed.
bed format support.
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.