Computational Biologist

  • Comments posted to this topic are about the item Computational Biologist

  • Steve, I would suspect that this gig is heavily using EAV

  • I am all in favour of bringing all options to the table but sincerely hope that we do not allow statistics to start to rule medical science. I have no problem with experimental science just as long as it is tempered on a theoretical basis combined with observational and evidential results. If that is the case then the best of luck to them (and therefore us).

    I have no scientific nor medical background so apologies for misuse of terms etc.

    Gaz

    -- Stop your grinnin' and drop your linen...they're everywhere!!!

  • The data may lead us one direction or another. But as we make decisions on that, we will get results, positive and negative, that will be refactored and qualify future decisions. I think it is a good thing. That's how medicine has developed, trial and error.

  • I find it interesting that the medical field has come full circle back to each body reacts differently. For a while it seemed medicine used the cookie cutter approach to medicine. While detailed data analysis can lead to better treatment the amount of data that needs to be analyzed is staggering. Medicine needs to keep in mind the key variance may not even be data that has been collected.

    My other thought is up to this point are there Computational Biologists?

  • Epidemiology

    So, this makes perfect sense. Lets take radiation treatments - and some of the variables.

    - what is the size of the daily treatment?

    - how long should it last, (one minute, two, etc.), along with the intensity.

    And, then how many days? 20?, 30?, 40?

    All in an effort to understand what has worked, what has not, and to find the most optimum approach. And the more data, the more accurate they can determine what is effective.

    This will only progress. In medicine, anything that has any sense of dosage, math will help to determine and improve what the correct value should be. - And in the future, it will start to vary more based on any given individual.

    The more you are prepared, the less you need it.

  • I love the fact that there are computational biologists, but there are a lot of common assumptions in the field of medicine that really need to be examined. Like studying men for cardiac issues and applying those findings to women when women have different symptoms for heat attack and are excluded or in the vast minority of the study population. Likewise, women have different uptake rates and should have different prescription dosages than men but frequently don't.

    -----
    [font="Arial"]Knowledge is of two kinds. We know a subject ourselves or we know where we can find information upon it. --Samuel Johnson[/font]

  • fregatepllada (3/15/2015)


    Steve, I would suspect that this gig is heavily using EAV

    Perhaps, but that's still just a storage design for the data. I bet there are some interesting queries written to try and analyze what works well and what shouldn't be tried.

  • Gary Varga (3/16/2015)


    I am all in favour of bringing all options to the table but sincerely hope that we do not allow statistics to start to rule medical science. I have no problem with experimental science just as long as it is tempered on a theoretical basis combined with observational and evidential results. If that is the case then the best of luck to them (and therefore us).

    I agree. I think this is a place where a large blend of human knowledge/decision making is tempered, not governed, by the data analysis.

  • Steve Jones - SSC Editor (3/16/2015)


    Gary Varga (3/16/2015)


    I am all in favour of bringing all options to the table but sincerely hope that we do not allow statistics to start to rule medical science. I have no problem with experimental science just as long as it is tempered on a theoretical basis combined with observational and evidential results. If that is the case then the best of luck to them (and therefore us).

    I agree. I think this is a place where a large blend of human knowledge/decision making is tempered, not governed, by the data analysis.

    Agreement here as well. Using the data for the options for treatment and the statistical variations on the markers in each cancer patient is a great thing, as long as there is not another parallel calculation as to if the cost of the treatment for each option fits inside the terms of the medical insurance held by the patient. We should use every tool possible to find treatment and restore health. Did we do all this just for money, or so that all this computer power can add to the quality of life?

    Not all gray hairs are Dinosaurs!

  • Gary Varga (3/16/2015)


    I am all in favour of bringing all options to the table but sincerely hope that we do not allow statistics to start to rule medical science.

    That barn door was opened a long time ago. Statistics has long dominated medical science. And to the benefit of everyone. Fewer wild-*** guesses, and more data driven validation.

    The more you are prepared, the less you need it.

  • It is quite odd to see your responses when working in Bioinformatics. Computational biology / bioinformatics started in the seventies, and statistics has always played a very important role in biology (I work in a biostatistics team).

    This field has a lot of interesting challenges on the dba side too: data is still just data but we have to deal with plenty of it.

    In fact we get overwhelmed with data, the hardest part is making sense of it πŸ™‚

  • eddy.b (3/18/2015)


    It is quite odd to see your responses when working in Bioinformatics. Computational biology / bioinformatics started in the seventies, and statistics has always played a very important role in biology (I work in a biostatistics team).

    This field has a lot of interesting challenges on the dba side too: data is still just data but we have to deal with plenty of it.

    In fact we get overwhelmed with data, the hardest part is making sense of it πŸ™‚

    Interesting. Never had heard of it before.

  • Gary Varga (3/16/2015)


    I am all in favour of bringing all options to the table but sincerely hope that we do not allow statistics to start to rule medical science. I have no problem with experimental science just as long as it is tempered on a theoretical basis combined with observational and evidential results. If that is the case then the best of luck to them (and therefore us).

    I have no scientific nor medical background so apologies for misuse of terms etc.

    The end quote addresses that pretty heavily in the article I think. β€œWe’re not the most important people,” he said, β€œbut we can help.”

    I work a lot with data scientist, mainly helping them with access to data and how we can better structure/manage data on the backend to help them run complex statistical models. We do this on consumer paths in digital advertising such as your journey through the digital marketing channels such as Google searches to social media channel clicks.

    When looking at those paths, you cannot assume every consumer path is the same or similar for every client. Different clients, different advertising, different products and services all alter the consumer's actions in some way. The same would obviously be true for patients who are different people who require different and more specialized treatment as they begin their journey to become healthy.

    If you can factor in dynamic variables such as the client on my end and the patient on their end, magic can happen. But, it would not be wise to think it's the one source of truth. It's just more quality intelligence based on data and science that help make better decisions among everything else.

    If you don't, which we did at first to automate that one-size-fits-all type of predictive modeling, bad things occur. That's obviously because not every client is the same much like not every patient is the same. These types of approaches greatly improve how we approach solving complex problems like curing cancer per patient rather than trying to cure cancer for the world if that makes sense. πŸ™‚

Viewing 14 posts - 1 through 13 (of 13 total)

You must be logged in to reply to this topic. Login to reply