• Question: What process do you follow to find answers?

    Asked by shreedesai to Lyn, Katy, Paul, PB, Ruth on 14 Jun 2013.
    • Photo: Ee Lyn Lim

      Ee Lyn Lim answered on 14 Jun 2013:

      To get an answer, first of all you have to ask a question! The trick is knowing what questions to ask. It’s almost like playing (a sensible version of) Twenty Questions. Say you want to know why Cell A produces more of Protein P when you treat it with Drug X – so you break it down to a series of yes/no questions – is it because Drug X makes Protein Q work faster, or does it actually block Protein R from doing what it should? Then you can go away and design experiments to test whether Protein Q and Protein R and working properly, and your results will answer either ‘yes’ or ‘no’ to those questions. Of course, once you’ve discovered that Drug X does indeed stop Protein R from working, you’ll have to ask why it does that…and the process repeats! Eventually you’ll have enough yes/no answers to piece together a full picture, and you’ll finally know what EXACTLY it is that Drug X does in Cell A!

    • Photo: Katy Brown

      Katy Brown answered on 14 Jun 2013:

      I mostly use a process called data mining. I take all the data I can find for a particular virus and a particular animal. I then perform a basic screen which gives me thousands of bits of DNA from the animal which look like they originated from the virus. I carry out lots of different types of analysis on these bits, until I only have the most interesting, best quality ones left. I then look at these one by one to see what they are and where they came from.

    • Photo: Peter Balfe

      Peter Balfe answered on 15 Jun 2013:

      We call it “the scientific method”. The idea that you set up a model in your head, “if this is true then that must happen, whereas if it isn’t something else will”. We call these hypotheses. The “null hypothesis” is a special case of this.

      For example my theory (which I’ve just made up) is that trees are taller in the middle of a wood than at the edges. First define what you mean by the middle, say half the area of the wood, with the other half being the edge. The null hypothesis is that all trees are on average the same size everywhere in the wood. Now go test! Measure some trees in both areas (which ones? how do you pick at random?), then compare their average heights. If you’re null hypothesis is right, the averages in both areas will be the same (more or less, fancy doing some statistics anyone?). If they’re not then you’ve rejected your null hypothesis, the experiment has worked! Of course your original theory could still be wrong – the trees might be shorter in the middle!

      Actually, having posed the question, – does anyone know the answer? I have no idea!

    • Photo: Paul Waines

      Paul Waines answered on 17 Jun 2013:

      First of all you need to have a clear question, or theory, in mind. This might come about as a result of a previous experiment- a lot of science is about asking the right questions!

      For example, I have a theory that my bacteria will grow better on plastic pipe rather than copper pipe.

      How do I test my theory? I do this by designing an experiment to prove (or disprove) it.

      I might do this by setting up lots of repeated tests e.g. lots of identical bits of pipe which are treated with exactly the same type and amount of bacteria, which are then allowed to grow. I do it this way because I’m trying to copy what I would expect to see in real life on a small scale, so I need to take small, natural variations (common in nature) into consideration.

      After I have counted my bacteria (there are lots of ways of doing this, so I have to think about what might be the best way for this experiment), I can then look at the results (data) and come up with a conclusion- has my theory been proved or disproved?

      depending on the results and my conclusion, I might have more questions (and more potential experiments) than I started with!

    • Photo: Ruth Mitchell

      Ruth Mitchell answered on 18 Jun 2013:

      The question you are asking will determine what process you follow: in terms of how the experiment is designed and also what aspect of the experiment you want to focus on.
      Personally I look a lot at protein levels of different messengers in the body when I treat cells with certain drugs to see how they react. But it is also interesting for me to look at RNA (the molecule that is then made into proteins).

Comments