Salmon v0.5.0

This release of Salmon brings some significant new features (hence the move from v0.4.X to v0.5). Note: The update to the indexing infrastructure means that you will have to re-build any existing indices (even if you choose to use the original FMD-based index), since the index version has changed and extra information is now recorded in the indices.

New Features

• Bootstrap estimates — You can now generate bootstrap samples to measure (technical) variance in the abundance estimates. This behavior is enabled by passing --numBootstraps k to the quant command. This will generate k bootstrap samples and place them in a file quant_bootstraps.sf. Bootstrap sampling is parallelized using the same number of threads requested for quantification, and though generating bootstraps does increase overall runtime, the process is very efficient. In addition to being generally useful for assessing confidence in the abundance estimates, these bootstraps allow Salmon to be used with a new class of differential expression tools.
• (Optional) Quasi-mapping based index — Salmon now supports two different types of indices. The FMD index, which is the index that Salmon has traditionally used up until this release, and a new suffix-array-based quasi-mapping index. When building the index, you now must specify what type of index you wish to build. The type of index is specified with the --type argument, and the options are --type fmd (the original index) or --type quasi (the new index). Support for both types of indices is being maintained in this release because the quasi-index is still mildly experimental (though seems to work very well in practice). You need only specify this option when building the index. Salmon will automatically determine how quantification should proceed based on the type of index with which it is run. The main benefit of the new quasi index type is that it speeds up mapping, sometimes significantly.

Salmon Beta v0.4.2

Fix bug (affecting only alignment-based Salmon), where a specific probability was not properly translated out of log-space.

Salmon Beta v0.4.1

This release incorporates a couple of bug fixes from v0.4.0 --- for full details about the newest features, see the release notes for that release.

Bug fixes

• Fixed an incompatibility between the new output format (i.e. dropping FPKM) and the bias correction driver.
• Fixed a bug in alignment-based Salmon, where the error model could be incorrectly applied to orphaned reads in a paired-end data file.

Salmon Beta v0.4.0

Important change to output format

The FPKM column of the output has been removed. This closes issue #3, as Richard has made a compelling case that FPKM should be completely deprecated from use and its inclusion in the output might promote undesirable practices in downstream analysis.

Fixes

This release also fixes some build issues that prevented v0.3.2 from compiling successfully under OSX using the default compiler and linker. These issues have been resolved, and v0.4.0 should build properly on OSX 10.10 (using the latest version of the system's compiler). Note: Only the official system compiler (Apple's clang) is supported on OSX; building with a non-native compiler (e.g. GCC) on OSX is still experimental.

Some new features and deprecations

This beta brings some important improvements and changes to Salmon. Certain program options (that previously affected the behavior of secondary streaming passes over the data) have been marked as deprecated and will be removed in a future release. The most notable of these is the -n (number of required observations) parameter. Salmon now optimizes the dataset until data-dependent convergence. This is possible as this release brings with it a novel hybrid optimization strategy that requires only a single streaming pass over the data (regardless of data set size). This leads to much faster run times on small datasets, and also, improved overall accuracy. The new hybrid optimization strategy is the default.

New command line flags

A new command-line flag, --useVBOpt has been added. This flag allows the user to toggle between the Variational Bayesian EM algorithm and the "standard" EM algorithm in the second phase of the hybrid optimization. Some previous literature has shown that the Variational Bayesian EM is often slightly more accurate, though we are currently testing both.

To close issue #2, it is now possible to specify the parameters of the fragment length distribution for single-end experiments. This is useful for correctly estimating the effective-length of the transcripts in single-end experiments. The mean of the fragment length distribution can be specified with --fldMean and the standard deviation with --fldSD.

Salmon Beta 0.3.2

New in this release

• Faster lightweight-alignment (requires rebuilding the index)
  • The index is now a dense rather than sparse suffix array-based index. This improves
    lightweight mapping speed 30 -- 50%.
• Modeling of non-uniform fragment start distribution
  • Salmon now learns (by default) a fragment position start distribution to account for systematic
    non-uniformity in transcript coverage. This should improve accuracy in datasets exhibiting this
    type of behavior. This feature is still experimental and can be disabled with the --noFragStartPosDist flag.

Latest successful Salmon build from Travis-CI

TravisCI

Update README.md