CStone paper: data for method S1

These datasets were simulated in order to demonstrate the effects of chimerism on differential expression analysis as described in supplementary method S1 of our manuscript titled “CStone: A de novo transcriptome assembler for short-read data that identifies non-chimeric contigs based on underlying graph structure.”. In brief from the 26,679 cDNA reference transcripts obtained from Ensembl, described for D. melanogaster (table 1 within manuscript), five read datasets, each containing two million read pairs, were simulated. For each, low level read count variation was introduced by allowing the number of reads required for an even coverage, as described in the manuscript, to be increased by a factor of between 0 and 0.2. The expected per site coverage provided by 2 million reads pairs across the 26,679 transcripts was slightly greater than 7X. These five replicates were allocated to condition A. A further five replicates were then simulated, where this time 1000 of the reference transcripts were selected at random for possible over-expression. For these 1000 genes the number of reads required was increased above the background variation by a factor of between 1 and 5. These five replicates were allocated to condition B. 

Related software to this project are:
1. CStone <
2. CSReadGen
3. CView
4. ChimSim
5. TVScript

A related poster discussing the the identification of chimerism during assembly is available here (DOI: 10.5281/zenodo.6022493) and one discussing the effects of chimerism is available here (DOI: 10.5281/zenodo.6023170).

General details of the project are available here.