1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Need Help in Understanding Genetic Classification of Breast Cancer Tumour

Discussion in 'Nottingham and Derby Medical School' started by half_pirate, Apr 23, 2012.

  1. half_pirate

    half_pirate New Member

    Joined:
    Apr 23, 2012
    Messages:
    4
    Likes Received:
    0
    Hello.

    I'm doing Modelling & Optimisation. So I'm not a medical student, applicant and anything related to medical field. But my Easter coursework has got to do with it. I have a slightly modified version of real-world medical data concerning genetic classification of breast cancer tumour. I attached a screenshot of the data so you can have a look and understand what I mean.

    In that screenshot, there's a figure. My question is, in that figure, there's a bunch of numbers in each row. I need to know what are those numbers represent basically? Like are they in a range from 0 to ??? where 0 is what? Does 0 mean safe? Normal? Healthy? Unknown? what about 30? 260? etc. Also at the top row consists header like er, pr, o2, o3,o7, please tell me what are those?

    Don't explain too details. Just gimme the basic definitions. I've thought I should at least understand the data set before analysing my statistics.

    Hope any of you to reply this a.s.a.p.

    Many thanks!
    half_pirate
     

    Attached Files:

  2. Martigan

    Martigan Super Moderator

    Joined:
    Mar 24, 2009
    Messages:
    2,535
    Likes Received:
    28
  3. half_pirate

    half_pirate New Member

    Joined:
    Apr 23, 2012
    Messages:
    4
    Likes Received:
    0
    Already had a look at the wikipedia. What's more important is what are those number basically represent. Still need help here for anyone out there. x
     
  4. House Jr.

    House Jr. New Member

    Joined:
    May 27, 2009
    Messages:
    407
    Likes Received:
    0
    Those numbers could simply be the number of cases with that specific cancer cell receptor and the subsequent class of the tumour upon surgical removal. What those classes mean though is unclear. You should really try to find the actual paper for this table.
     
  5. half_pirate

    half_pirate New Member

    Joined:
    Apr 23, 2012
    Messages:
    4
    Likes Received:
    0
    Thanks for replying that helps a lot but you said 'Those numbers could simply be..'

    Are you absolutely sure about that or was that an assumption?

    In the screenshot it stated that 'The attributes are genetic markers that have been determined by assessment of the tumour and the class...'

    I just to make sure what are genetic markers? Are they really like you said - 'number of cases with that specific cancer cell...'?

    Anyone else who knows about this are welcome. Still need help.

    Cheers.
     
  6. House Jr.

    House Jr. New Member

    Joined:
    May 27, 2009
    Messages:
    407
    Likes Received:
    0
    I'm afraid it's only an assumption, but after reading it again it seems more likely that the attributes are referring to the location of a DNA sequence (the genetic marker) rather than the receptors themselves. So the numbers could refer to the number of cases with the DNA sequence in the column and the tumour class in the row. Maybe someone with more knowledge will correct me though, sorry I couldn't be of more help.
     
  7. half_pirate

    half_pirate New Member

    Joined:
    Apr 23, 2012
    Messages:
    4
    Likes Received:
    0
    Yes, I hope there's someone out there will read and reply this sooner.
    But I really thank you for your replies though. Definitely give some idea to narrow down my research meanwhile.
     
  8. terrified_tony

    terrified_tony New Member

    Joined:
    Sep 1, 2009
    Messages:
    180
    Likes Received:
    0
    The data file he has attached is a snapshot, not the full data shot, so each line represents a patient, therefore the numbers relate specifically to that genetic marker for that case. It would make sense that the number is related to the expression of that receptor, how the two are related - linearly or not is irrelevant to the OP as it is a data processing task and not a task of understanding cancer.

    On the page - G52IMO - Introduction to Modelling and Optimisation - 2011/12 , where this project comes from there is a question regarding the numbers in the data...

    "What do the numbers in the data represent (are they some form of 'intensity')? The numbers represent a form of protein intensity, but the details are completely unimportant for you. You are just analysing the data from a 'pure' perspective - what can you learn from exploring the numbers (and existing class labels)?"



    I realised this was an old question only after i had answered it! DOH!
     
  9. Poe

    Poe New Member

    Joined:
    Oct 29, 2012
    Messages:
    1
    Likes Received:
    0
    I know it's an old question but just in case ...

    The dataset shows the genetic markers for 15 patients with breast cancer and, based on these, which of the main breast cancer types their disease falls into (see last column) ... Breast cancer is actually at least 4 main types - Basal like (or triple negative), two luminal types and a HER2 enriched/amplified type. Eventually there will hopefully be drug targets for each type and sub-type.

    Comprehensive molecular portraits of human breast tumours : Nature : Nature Publishing Group

    The researchers have looked at the genomes of patients and undertaken cluster analysis (a type of multivariate statistics) to look for broad clusters/ types ... It's not the exact same research as the extract but they have some nice graphics to show their results ...
     

Share This Page