I'm just beginning to wade into the world that is R. I'm currently having some very basic problems. The problem is so basic that I cannot find any examples out there. I don't have any problem getting my data into R, but what I want to know is how to best group the data prior to importing.
Please refer to the spreadsheet for details. Should I use option_1 (sheet 1) or option_2 (sheet 2) as the format for my data? Does it matter? Will this affect what I can do for my analysis?
What I ultimately want to do is compare the data (ANOVA) from the SC animals to the SD animals. There are some data points missing for some of the data as you can see from the file. I want to be able to compare the data using either the protein column or the peptide column. Do I need the unique_ID column?
The actual data has more samples and more data points.
Any and all suggestions are greatly appreciated.
Hi Peter,
you may check the iris data set within R as an example just by typing iris into the prompt of R and pressing enter.
To get more information on this data set type
> ?iris
However, I recommand you to check some books on R such as RinAction or Rbook or use some freely available tutorials that leads you to topics such as multiple comparisons of the means of an ANOVA model or checking for equal variances and more.
Hope it helps.
Boran
I do not understand Tims answer... can it be that there is a typo?
[EDIT] I think I understood. But I am still puzzled because SC1 ~ SD1 would give the (linear) regression SC1 vs. SD1, what seems not related to the question. [\EDIT]
I vote for option 2, what is called the "long format". I would change the column-name for the values (column "D") from "SC1" to "Value" or something.
Generally, a "good" data layout is to have one measurement per row, and one variable per column. In your "option 2"-sheet, the response variable is in "D", and the group variable is in "E" (and other covariables like combined cells, protein, and peptide in the columns "A" to "C").
However, it is possible to convert the data (in R) for the "wide" to the "long" format and vice versa, so it actually does not matter much (see reshape(); it's a little difficult to grasp, so having the data in straight-away in the ideal layout saves time and nerves).
What puzzles me more is:
- will you do the comparison seperately for each peptide or will you pool all peptides (and if so, would this be sensible)?
- Your group variable has 4 levels ("SC1","SC2","SD1","SD2"). Are these really 4 distinct groups or do you actually have the two groups "SC" and "SD", both measured in duplicates? If so, you may (if this information is important) provide the "number of replicate" in another column, but in the column "E" you must have either "SC" or "SD".
If you have the data in the long format you can quickly split the data by subgroups using split(), apply some calculation to subgroups by tapply() (of by lapply() to the list you get after split()) and so on. Sometimes the functions stack() and unstack() are helpful.
Given you have the data in the long format in x and you want to perform an ANOVA comparing the values of "SC1" (in column "D") between the four groups (as specified in "Data_grp") you can simply call
aov(SC1~Data_grp, x)
Now you may understand why I suggested to rename the columns "SC1" to "Value". Then it would read:
aov(Value~Data_grp, x)
Please also consider what Boran said: make sure that ANOVA is the right tool answering the question you have and that the assumptions are not heavily violated. Check the residuals. For instance, plot(aov(SC1~Data_grp, x)) shows that the residuals in this model are far from being normal and homoscedastic, so the results from the ANOVA can't be sensibly interpreted.
Hi all,
thanks for the responses. I was hoping for a shortcut to get my analysis moving, but clearly I'll need to go do some more R homework.
To answer your questions Jochen;
I have four distinct groups, each with an individual measurement for the peptides. I want to compare the data at either the protein or peptide level. For some proteins, I have multiple peptides (each with distinct measurements, ie. 1 value from each animal). Where this is the case, I want to combine this data into one comparison of SC vs SD.
Where I have only one peptide for a given protein, I will only be using the single peptide for comparison.
Thanks again,
Peter
Hi Peter
"Thanks for the responses. I was hoping for a shortcut to get my analysis moving, but clearly I'll need to go do some more R homework."
There are some great free texts on the R-website that guide you through everything slowly (really great as reference material as well). Additionally Andy Field's book is amazing and will guide you through any basic analyses you need. Furthermore, try these resources for quick hints / help as in R you can get stuck on minor coding issues which they might resolve:
http://rprogramming.net/ - General help and topics
http://www.statmethods.net/ - great online guide and quick help options
Additionally, I would recommend downloading RStudio which will drastically increase its ease of use and power.
https://www.rstudio.com
If you're comparing SC1 between groups (Data_grp) , use option two.
aov(SC1 ~ Data_grp, data=your data)
- Badges
- Science topic