collapseTranscripts: cluster a gene's transcripts and calculate cluster-level...

In ballgown: Flexible, isoform-level differential expression analysis

Description

cluster a gene's transcripts and calculate cluster-level expression

Usage

collapseTranscripts(gene, gown, meas = "FPKM", method = c("hclust",
  "kmeans"), k = NULL)

Arguments

gene	which gene's transcripts should be clustered
gown	ballgown object
meas	which transcript-level expression measurement to use ('cov', average per-base coverage, or 'FPKM')
method	which clustering method to use: 'hclust' (hierarchical clustering) or 'kmeans' (k-means clustering).
k	how many clusters to use.

Value

list with two elements:

• tab, a cluster-by-sample table of expression measurements (meas, either cov or FPKM), where the expression measurement for each cluster is the mean (for 'cov') or aggregate (for 'FPKM', as in gexpr) expression measurement for all the transcripts in that cluster. This table can be used as the gowntable argument to stattest, if differential expression results for transcript *clusters* are desired.

• cl output from clusterTranscripts that was run to produce tab, for reference. Cluster IDs in the cluster component correspond to row names of tab

Author(s)

Alyssa Frazee

See Also

hclust, kmeans, clusterTranscripts, plotLatentTranscripts

Examples

data(bg)
collapseTranscripts(bg, gene='XLOC_000454', meas='FPKM', method='kmeans')

ballgown documentation built on Nov. 8, 2020, 5:46 p.m.