Which Platform Should I Use?
Cofactor offers two of the three Next-Generation sequencing platforms: the Applied Biosystems SOLiD and the Solexa/Illumina. We can arrange for runs on the Roche/454 Titanium. Each platform has strengths and weaknesses that make it more or less appropriate for your application. At Cofactor, platform choice and optimization is our job, you don’t have to worry! You describe your research question to us, and then we design your project to choose an optimal platform (or combination of platforms) that maximizes your bottom line.
Applied Biosystems SOLiD 4
The Applied Biosystems SOLiD 3 has the shortest but also the highest quantity of reads. The SOLiD produces up to 500 million 50×50 or 50×25 bp reads per slide. 500 million 50bp single end reads per slide. As with the Illumina, Mate-Pairs produce double the output by duplicating the read length on each end, and the SOLiD supports a variety of insert lengths like the 454. The SOLiD can also run 2 slides at once to again double the output. SOLiD has the lowest *raw* base qualities but the highest processed base qualities when using a reference due to its 2-base encoding. Because of the number of reads and more advanced library types, we recommend the SOLiD for all RNA and bisulfite sequencing projects.
Solexa/Illumina
The Solexa/Illumina generates shorter reads at 36-75bp but produces up to 160 million reads per run. All reads are of similar length. The Illumina has the highest *raw* quality scores and its errors are mostly base substitutions. Paired-end reads with ~200 bp inserts are possible with high efficiency and double the output of the machine by duplicating the read length on each end. Paired-end Illumina reads are suitable for de novo assemblies, especially in combination with 454. The large number of reads makes the Illumina appropriate for de novo transcriptome studies with simultaneous discovery and quantification of RNAs at qRT-PCR accuracy.
Roche/454 FLX
The Roche/454 FLX with Titanium chemistry generates the longest reads (350-500bp) and the most contiguous assemblies, can phase SNPs or other features into blocks, and has the shortest run times. However, 454 also produces the fewest total reads (~1 million) at the highest cost per base. Read lengths are variable. Errors occur mostly at the ends of long same-nucleotide stretches. Libraries can be constructed with many insert sizes (8kb - 20kb) but at half of the read length for each end and with low efficiency.
