We performed RNA-Seq of in vivo bacterial samples and identified some key up- and down-regulated pathways.

We compared bacteria during infection with conventional agar plate.

Which could be the next steps to validate our findings? We are interested in antibiotic and vaccine developemnt.

  • $\begingroup$ You're going to need to be a lot more specific. To start with, when you say a pathway is up or down-regulated, what comparison are you making? Is there a treatment and you're comparing treated versus control? Are there different populations and you're comparing to each other? $\endgroup$ May 11, 2015 at 14:20
  • $\begingroup$ Updated. We compared bacteria during infection with conventional agar plate. $\endgroup$
    – biotech
    May 11, 2015 at 15:08

1 Answer 1


Okay, first, I would say to everyone planning a high-throughput experiment: Hypothesis first, then experiment. Otherwise you're setting up an expensive and time-consuming test that's not going to answer the questions you want answered.

In this case, the question your experiment is designed to answer is "What genes are crucial to growth in vivo that are not crucial to growth on a plate?", which is probably not the clearest model for vaccine/antibiotic design.

I guess you could say that a difference between the plate sample and the in vivo sample is the presence of a host, and therefore it's possible that upregulated pathways represent the bacteria's host adaptation, and maybe you can attack the bacteria by attacking its ability to colonize the host. A gene involved in immune evasion could make a good target.

(That's a very messy model, as there are tons of other differences going on, but since you've already run the experiment it's probably the best you've got.)

So then, use your data for a new round of hypothesis generation. Take your pathways and come up with a way that they might be allowing the bacteria to adapt to the host-- maybe a secretory pathway that could be interfering with the target immune system, or some membrane signalling molecules that get upregulated, or whatever.

Now you have a hypothesis, so design an experiment. For instance, take a key upregulated member of a pathway. The hypothesis is that it's crucial to growth in a host but not to growth in general. So, knock it down, either with chemical inhibitors or with genomic editing, and see if it's less effective at host colonization.

(Although host colonization effectiveness is pretty difficult to measure ethically if your target organism is humans. But you have some sort of proxy or model system to measure that, right?)

  • $\begingroup$ Thanks for your helpful answer @Transcriptase. We work with a pig model of infection, since is a swine pathogen. $\endgroup$
    – biotech
    May 12, 2015 at 6:59
  • $\begingroup$ Why do you think this is not the clearest model for vaccine/antibiotic design? Do you know better alternatives? $\endgroup$
    – biotech
    May 12, 2015 at 7:01
  • $\begingroup$ Okay, that does make experimental design much easier. $\endgroup$ May 12, 2015 at 13:16
  • $\begingroup$ The reason I say it's not the clearest model, especially for antibiotic design, is that you're investigating the difference between bacteria in the host and bacteria on the plate. An antibiotic in particular probably isn't going to care about that-- it should kill in both situations, right? And for a vaccine, there are so many differences between culture and in vivo that I'm not sure that you can attribute differences you see to immune evasion. Are there immune-compromised pig models? In vivo samples from those would be a much cleaner control for immune evasion. $\endgroup$ May 12, 2015 at 13:28
  • $\begingroup$ Your comment contains some assertions that I agree or I had not think about it. Others are more objectionable. First, antibiotic resistance would be lower in vivo, due to reduced metabolic rate and that common antibiotics rely on this assumption. $\endgroup$
    – biotech
    May 13, 2015 at 9:24

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .