You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: doc/source/salmon.rst
+16-2Lines changed: 16 additions & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -269,8 +269,8 @@ mode, and a description of each, run ``salmon quant --help-alignment``.
269
269
header sections must be identical).
270
270
271
271
272
-
Description of important options
273
-
--------------------------------
272
+
Description of some important options
273
+
-------------------------------------
274
274
275
275
Salmon exposes a number of useful optional command-line parameters to the user.
276
276
The particularly important ones are explained here, but you can always run
@@ -296,6 +296,20 @@ by RSEM), but using the default scoring scheme and allowing both mismatches and
296
296
indels in alignments. These setting essentially disallow indels in the resulting
297
297
alignments.
298
298
299
+
""""""""""""""""""""""""""""""
300
+
``--meta``
301
+
""""""""""""""""""""""""""""""
302
+
303
+
As with the flags described above, this is a "meta-flag" that simply enables some options
304
+
that may make more sense when quantifying metagenomic data. Specifically, the ``--meta``
305
+
flag sets the following options:
306
+
307
+
* The abundance optimization is initialized from the uniform distribution (compared to the default of using a weighted combination of the uniform intialization and the abundances learned during the online optimization)
308
+
309
+
* Rich equivalence classes are disabled. Using rich equivalence classes with metagenomic data should not be particularly problematic, but since they have been developed and tested most in the context of bulk RNA-seq quantification, they are currently disabled under this flag.
310
+
311
+
* The EM algorithm is used for abundance optimization instead of the default VBEM optimization. Neither is universally better than the other, but the parameters for the VBEM (e.g. the prior size and type) are set based on typical bulk RNA-seq transcriptome samples, and so may be less appropriate in the metagenomic context. Hence the ``--meta`` flags opts for the basic EM algorithm instead.
0 commit comments