I'm a newbie to the world of NGS, but I thought it might be a suitable and cheaper alternative to doing traditional Sanger sequencing for this particular application. However, I'm not so sure on how to approach calculating this cost, so I was hoping to reach out and see if someone could help here. This is for my grad school research project.

Problem Statement

I am about to make a rational library of ~1000 protein complexes belonging to the same family of proteins. There are 4 protein's plasmids to make per complex, and they are to be made by a combination of synthetic genomic cloning methods using cDNA, and multi-site directed mutagenesis on existing plasmids. Before testing their function as a complex, I need to verify that the cloning steps succeeded by sequence verifying each plasmid.

I could do traditional Sanger sequencing on each of the protein sequences, but given that each protein is ~2kb long, I would need 4 single direction sequencing reactions per plasmid to cover the CDS. In total, per protein complex, that's 16 sequencing reactions. At USD5 per sequencing reaction, each protein complex would cost USD80 to sequence. At 1000 sequences, that would be USD80,000 in Sanger sequencing.

I was thinking of doing something else - after creating the library, say in chunks of 384, send them for NGS on an Illumina MiSeq.

Request for Help

Given that:

  • I am not wedded to any particular NGS protocol,
  • Our only goal here is to, by one way or another, sequence verify the cloned plasmids.
  • 1000 proteins is the right ballpark for the number of proteins that need to be sequence verified. 4 plasmids per protein complex.

What would be the appropriate sequencing strategy that could help us get way below 80K for sequencing costs?

  • $\begingroup$ How would NGS help you? Are you planning on mixing all 1000 plasmids together and trying to get complete coverage on all of them? $\endgroup$ – MattDMo Sep 17 '15 at 2:53
  • $\begingroup$ @MattDMo: Yes, that's the plan. It's 4 plasmids per sample, 1000 samples in total, each needs to have (at minimum) its 2 kb insert sequenced. But if we could get full coverage of the plasmid that's cool too. $\endgroup$ – ericmjl Sep 17 '15 at 15:20

Most NGS approaches should enable you to sequence all those plasmids at very low cost-per-base values (Please see here for more on prices). The challenge in sequencing them all, if there are sequences common to all your plasmids, will be to get adapters for each sample/plasmid so that reads can be mapped back to the correct plasmids.

  • $\begingroup$ Thanks Ankur - yes. These are a family of polymerases with 4 proteins. We are planning to make a rationally designed library of 1000 polymerase variants. 4 plasmids per sample, each of a different sequence, 1000 samples in total. $\endgroup$ – ericmjl Sep 17 '15 at 15:22
  • $\begingroup$ OK - as long as there are no sequences that overlap between those plasmids you should be able to get good depth of sequence for a grand or two I would imagine and then map back to the reference sequences to see if assembly has worked. NGS is defo the way to go. $\endgroup$ – Ankur Chakravarthy Sep 17 '15 at 15:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.