Description

The Quake lab used single cell RNA sequencing on 466 cells to capture the cellular complexity of the adult and fetal human brain at a whole transcriptome level. Healthy adult temporal lobe tissue was obtained from epileptic patients during temporal lobectomy for medically refractory seizures. The Quake lab was able to classify individual cells into all of the major neuronal, glial, and vascular cell types in the brain. For this track only the adult cells were considered, resulting in a 331 cell subset.

Display Conventions

In Full and Pack display modes, expression for each cell type is represented by a colored bargraph, where the height of each bar represents the median expression level across all samples for a cell type, and the bar color indicates the cell type.
     
The bargraph display has the same width and cell type order for all genes or transcripts. Mouse hover over a bar will show the cell type and median expression level. The Squish display mode draws a rectangle for each gene or transcript, colored to indicate the cell type with highest expression level if it contributes more than 10% to the overall expression (and colored black if no cell type predominates). In Dense mode, the darkness of the grayscale rectangle displayed for the gene or transcript reflects the total median expression level across all cell types.

Click-through on a graph displays a boxplot of expression level quartiles with outliers, per cell type. The track configuration page provides controls to limit the genes or transcripts and cell types displayed, and to select raw or log transformed expression level display.

The groups are colored according to the biased groupings in sections B and D of Figure 1 in the accompanying paper (seen below).
     

Methods

The raw sequences were downloaded using the SRA toolkit and an accession list from the SRA run selector. The reads were quantified to the hg38/GRCh38 reference using kallisto and a transcriptome file. The transcriptome file was generated from the HG38 reference genome and Gencode v23 comprehensive CHR annotation file5. The reference genome did not contain alternative sequences and had the overlapping genes from the PAR locus removed (chrY:10,000-2,781,479 and chrY:56,887,902-57,217,415) due to duplicate gene conflicts with the X chromosome.

Read quantification was performed at UCSC by the Genome Browser team. The resulting kallisto files were combined to generate both a transcript per million (tpm) and a gene per million (gpm) expression matrices using the UCSC tool kallistoToMatrix. A meta data file mapping each sample to its cell type was obtained from the A SOFT formatted meta data file was downloaded from the GEO ftp server. This file was then converted into a tab separated file which was used to map each sample to the Quake assigned cell type. Coordinates for the transcripts and genes were identified using the UCSC table browser and used to generate two bed6+1 file which map the ensemble gene/transcript names to chromosomes, genomic coordinates, strand, score and hugo gene names.

The meta data, expression matrix and bed6+1 coordinate files were passed into the UCSC tool expMatrixToBarchartBed. This script calculated the average expression value for each cell type and generated a bed6+5 file that is the base of the track. Finally the UCSC tool bedToBigBed was used to convert to a bigBed file for displaying.

Credits

Samples were produced and sequenced by the Quake lab. The raw fastq files for their RNA-seq are available online at the Sequence Read Archive.

The meta data, including the Quake lab assigned cell types, was obtained online through the GEO ftp server.

The read quantification and track generation was done by Chris Eisenhart at the UCSC genome browser group.

References

Spyros Darmanis, Steven A. Sloan, Ye Zhang, Martin Enge, Christine Caneda, Lawrence M. Shuer, Melanie G. Hayden Gephart, Ben A. Barres, and Stephen R. Quake. A survey of human brain transcriptome diversity at the single cell level. Biological Sciences - Neuroscience. 2015 May 18, 2015 112 (23) 7285-7290. PMID: 26060301; PMC: PMC4466750