* * *
-<h2 id="introduction">Introduction</h2>
+## Introduction <a name="introduction"></a>
RSEM is a software package for estimating gene and isoform expression
levels from RNA-Seq data. The new RSEM package (rsem-1.x) provides an
addition, it provides posterior mean and 95% credibility interval
estimates for expression levels.
-<h2 id="compilation">Compilation & Installation</h2>
+## Compilation & Installation <a name="compilation"></a>
To compile RSEM, simply run
Refer to the [UCSC custom track help page](http://genome.ucsc.edu/goldenPath/help/customTrack.html).
-<h2 id="example">Example</h2>
+## Example <a name="example"></a>
Suppose we download the mouse genome from UCSC Genome Browser. We will
use a reference_name of 'mm9'. We have a FASTQ-formatted file,
rsem-calculate-expression --bowtie-path /sw/bowtie --phred64-quals --fragment-length-mean 150.0 --fragment-length-sd 35.0 -p 8 --out-bam --calc-ci --memory-allocate 1024 /data/mmliver.fq /ref/mm9 mmliver_single_quals
rsem-bam2wig mmliver_single_quals.sorted.bam mmliver_single_quals.sorted.wig mmliver_single_quals
-<h2 id="simulation">Simulation</h2>
+## Simulation <a name="simulation"></a>
### Usage:
output_name.sim.isoforms.results, output_name.sim.genes.results : Results estimated based on sample values.
-<h2 id="acknowledgements">Acknowledgements</h2>
+## Acknowledgements <a name="acknowledgements"></a>
RSEM uses randomc.h and mersenne.cpp from
<http://lxnt.info/rng/randomc.htm> for random number generation. RSEM
also uses the [Boost C++](http://www.boost.org) and
[samtools](http://samtools.sourceforge.net) libraries.
-<h2 id="license">License</h2>
+## License <a name="license"></a>
RSEM is licensed under the [GNU General Public License v3](http://www.gnu.org/licenses/gpl-3.0.html).