+ Reply to Thread
Results 1 to 7 of 7

Thread: Display range from discrete data

  1. #1
    Points: 18, Level: 1
    Level completed: 35%, Points required for next Level: 32

    Posts
    2
    Thanks
    2
    Thanked 0 Times in 0 Posts

    Display range from discrete data




    Quick statistical/mathematical question. I have generated data on lethal dosages in several of my experiments. I would like to create a chart showing these lethal doses. The issue is that there is "space" between the lethal dose value and the next lowest dosage tested.

    Ex: LD = 2.0uM, next lowest amount tested was 1.5uM.

    From the above, how can I display the range that the lethal dose falls in? For the above example, I know that the lethal dose falls somewhere >1.5 but <=2.0.

  2. #2
    Super Moderator
    Points: 31,766, Level: 100
    Level completed: 0%, Points required for next Level: 0
    bugman's Avatar
    Posts
    2,255
    Thanks
    290
    Thanked 324 Times in 265 Posts

    Re: Display range from discrete data

    Its pretty common. Have a look at this paper that deals with censored data and dose response curves.

    http://www.pnas.org/content/75/9/4087.full.pdf
    The earth is round: P<0.05

  3. The Following User Says Thank You to bugman For This Useful Post:

    mas_snp (09-18-2013)

  4. #3
    Human
    Points: 12,676, Level: 73
    Level completed: 57%, Points required for next Level: 174
    Awards:
    Master Tagger
    GretaGarbo's Avatar
    Posts
    1,362
    Thanks
    455
    Thanked 462 Times in 402 Posts

    Re: Display range from discrete data

    Or simply run a logit model = logistic regression where 1 indicates “dead” and 0 “alive”.

    Also google at “LD50”. Maybe you want to pick out p=50 and solve what poison level that corresponds to. Quite often log-dose (instead of dose) is used as explanatory variable.

    - - -

    When someone writes “Quick question” does that mean that it was quickly written, thus carelessly written? Or does it mean that the writer demand a quick answer? I think that both are inappropriate.

  5. The Following User Says Thank You to GretaGarbo For This Useful Post:

    mas_snp (09-18-2013)

  6. #4
    Super Moderator
    Points: 31,766, Level: 100
    Level completed: 0%, Points required for next Level: 0
    bugman's Avatar
    Posts
    2,255
    Thanks
    290
    Thanked 324 Times in 265 Posts

    Re: Display range from discrete data

    This is a pretty simple method.

    And I think the question was more about display than analysis.
    The earth is round: P<0.05

  7. #5
    Super Moderator
    Points: 31,766, Level: 100
    Level completed: 0%, Points required for next Level: 0
    bugman's Avatar
    Posts
    2,255
    Thanks
    290
    Thanked 324 Times in 265 Posts

    Re: Display range from discrete data

    Quote Originally Posted by GretaGarbo View Post
    When someone writes “Quick question” does that mean that it was quickly written, thus carelessly written? Or does it mean that the writer demand a quick answer? I think that both are inappropriate.
    I usually give people the benefit of the doubt and treat it as a sort of sign of repsect, meaning "you guys probably know more about this than me, so it shouldn't take you long to answer..."

    I use it sometimes at work when I am talking to hydrologists. Its a term of phase generally, nothing more, nothing less. People who use text speak in their posts and CAPITAL LETTERS TO EMPHASIS THE FACT THEY ARE DESPERATE FOR HELP, are usually more careless with their posts imho, lol.
    The earth is round: P<0.05

  8. #6
    Human
    Points: 12,676, Level: 73
    Level completed: 57%, Points required for next Level: 174
    Awards:
    Master Tagger
    GretaGarbo's Avatar
    Posts
    1,362
    Thanks
    455
    Thanked 462 Times in 402 Posts

    Re: Display range from discrete data

    Quote Originally Posted by bugman View Post
    This is a pretty simple method.
    Yeah OK, logit is a pretty simple method. But in my view it is a good model.

    Quote Originally Posted by bugman View Post
    And I think the question was more about display than analysis.
    And I suggested to simply draw the s-shaped dose-response curve.

    P = 1/(1+ exp(-(a+b*x))) , where p is proportion dead and x is dose and a, b are parameters to be estimated.

    (And to mark the LD50-point.)

    - - - -


    Quote Originally Posted by bugman View Post
    Its a term of phase generally, nothing more, nothing less.
    We non-native English readers note that we don't always understand. And the native English writers can learn that that they are sometimes misunderstood. I have thought many times that: “OK, you have written a quick question, and I am going to give you a quick answer. None at all!”

  9. #7
    Points: 18, Level: 1
    Level completed: 35%, Points required for next Level: 32

    Posts
    2
    Thanks
    2
    Thanked 0 Times in 0 Posts

    Re: Display range from discrete data


    Thank you both for your responses. I believe what I am dealing with is a limited sample pool (all my data show "complete death" at 2.0uM) and a limited scope of observation (at 1.5uM, cells are alive). I don't have any data as "to what degree they are alive", just that they are not completely killed off. I'm not quite sure if I can do much statistical modeling with this data. Rather, I'd just like to be able to indicate in a chart that the lethal dose was 2.0uM, but the true lethal dose may lie anywhere between 1.5 and 2.0 (but above 1.5uM since cells survived in 1.5uM). Would it be appropriate to list the value as 1.75(+/-0.25)?

    ---
    And yes bugman, I used the term "quick question" to indicate that this was likely a simple question for educated statisticians. My field of study is outside statistics, so I'm sure my query is quite novice.

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats