Analysis of Differential Gene Expression in a Myotonic

Download Report

Transcript Analysis of Differential Gene Expression in a Myotonic

Analysis of Differential Gene
Expression in a Myotonic Dystrophy
Tissue Cultue Model
Matthew Tanner
Berglund Lab, Institute of Molecular Biology
CIS 407
Dec. 2nd, 2013
Myotonic Dystrophy 1 (DM1)
Symptoms
Multisystemic – affects skeletal
and smooth muscle (myotonia,
atrophy), eyes (cataracts), heart,
and endocrine system.
Genetic Pathology
DM1 caused by CTG
trinucleotide repeat expansion
in 3’ UTR of DMPK gene.
Unaffected: less than 34 repeats
DM1: 50 to >1000 repeats
Severity of disease increases
and age of onset decreases with
increasing repeat length.
Potential for small molecules to
alleviate molecular symptoms of DM1
Pentamidine
Anti-fungal, anti-protozoan drug.
Binds minor groove of DNA.
Coonrod, et al. (2013)
Wilson et al.,
2008.
Tissue Culture Model and Experiment
DMPK960
DMPK minigene
containing 960 CTG
repeats. Induces DM1
disease state.
Transfect
into HeLa
cells. Drug 6
hours
afterwards.
24 hours after
transfection: harvest
whole-cell RNA.
Illumina
highthroughput
sequencing library
preparation (several
steps)
Illumina Hi-Seq 2000
massively parallel
sequencer
RNA-Seq Library Preparation
Pease & Sooknanan,
2012.
Illumina Raw Output
Four lines for each sequence:
Coordinates of read
Sequence
+
ASCII quality score for each base call (Phred-33)
Alignment and Analysis with Tuxedo Suite
Tophat 2.0 – align FASTQ reads that were cleaned up with Stacks’
process_shortreads to human genome.
Pass $SAMPLE variable in at command line:
qsub tophat.sh –N align0A –V SAMPLE=“0A”
Alignment and Analysis with Tuxedo Suite
Tophat 2.0 – align FASTQ reads that were cleaned up with Stacks’
process_shortreads to human genome.
Cufflinks – take mapped reads (accepted_hits.bam) and generate
transcript model of reads.
Cuffmerge – take individual transcript models (transcripts.gtf)and
merge into master transcriptome.
Cuffdiff – take mapped reads from individual treatments
(accepted_hits.bam) and, with aid of master transcriptome, compare each
sample pair-wise.
Take various Cuffdiff output files (differential gene expression, splicing,
promoter usage, isoforms, etc.) and analyze with original scripts or explore with
existing programs.
Integrative Genome Viewer
Integrative Genome Viewer
Visualization of differential
gene expression
log10 (FPKM + 1) of genes
at each dosage that are
associated the p53
network.
FPKM: fragments per
kilobase of exon model
per million mapped
fragments
high
med
low
none
Visualization of differential
gene expression
log10 (FPKM + 1) of genes at
each dosage that are
associated with the gene
ontology “Regulation of RNA
splicing” (GO:0043484)
What are
these
genes?
How about all the (significantly
differentially expressed) genes?
log2(foldchange) of 1210
genes with q < 0.05
between none and low
pentamidine dosage
A closer look
none
low
med
high
Gene ontologies found in this subset
Next steps?
• Analyze this and other clustered subsets by:
– Sequence analysis (motifs)
– GC/AT content
– Gene ontology enrichment