Skip to content

Latest commit

 

History

History
55 lines (42 loc) · 2.22 KB

README.md

File metadata and controls

55 lines (42 loc) · 2.22 KB
Raw

binner

binner is an R package for processing DNA fingerprint data. At present, it provides a complete workflow for AFLP analysis:

  1. reading ABI .fsa files
  2. normalizing electropherograms
  3. identifying and sizing peaks
  4. viewing individual electropherograms
  5. dropping, adding and renaming samples
  6. Automated peak-binning, using the RawGeno algorithm
  7. Generating presence-absence matrices for further analysis in R (or export for use in other programs, if you like)

Installation

You can install binner using Hadley Wickham's devtools package.

install.packages("devtools")
install_github("plantarum/binner")

Help

There is no vignette for the package. I have provided a reasonably complete example in the help file for the function readFSA. Open it from with R with ?readFSA (after you've loaded binner of course).

Alternatives

There are two R-based alternatives you may like to consider as well:

RawGeno

RawGeno currently requires the use of a second program, PeakScanner, to read the raw fsa data. binner is a bit slower than PeakScanner for reading fsa files, but implements the same sizing algorithm (local southern). Parts of binner may make their way into the next version of RawGeno, and the current version of binner has pilfered some key bits of RawGeno (most notably the binning algorithm).

If working entirely in R is useful to you, binner is currently preferable to RawGeno. binner also provides a GUI for checking and editing your bins. Beyond that it's a matter of preference.

AFLP

AFLP provides a more sophisticated workflow than binner or RawGeno, and provides a very powerful system for controlling for variation between capillaries, sequencers, sequencing runs etc. However, it seems to use a different sizing system, making it hard to make direct comparisons between electropherograms processed with AFLP and those using the PeakScanner/binner approach.