Doing multiple ttests within a dataframe. - r

Question

I would like to do a series of ttests on a dataset:
specifically, i'd like to do seperate ttest for rats 1-5 vs 6-10 for every gene.
I've tried to do this:
>goi2 <- (goi[-1])
control <- goi2[1:5,]
stress <- goi2[6:10,]
for (i in 1:92){
x <- control[,i]
y <- stress[,i]
x <= t.test(x, y)
# print(x=i)
}
but i get this error:
Error: Can't use matrix or array for column indexing
I've tried a few varieties of this but cant figure out why this wont work.
Im a complete newb to R, but not programming in general.
Dataset:
Gene,Rat_1,Rat_2,Rat_3,Rat_4,Rat_5,Rat_6,Rat_7,Rat_8,Rat_9,Rat_10
Oprd1,2.746,1.387,2.25,3.363,3.191,2.432,1.985,1.75,2.752,1.771
Grin2a,3.134,2.644,2.962,5.168,2.484,3.54,2.596,1.535,3.197,2.232
Grin2d(2),4.496,5.528,2.631,4.684,3.934,6.047,0.98,0.077,4.381,2.327
Oprm1,1.998,1.804,1.611,1.712,3.672,3.215,0.249,1.248,1.758,2.671
Scn2b,137.35,97.158,113.65,141.93,77.295,133.02,88.872,75.586,108.96,97.626
Ntf3,0.989,1.835,1.604,1.133,0.889,0.782,0.918,2.241,2.216,3.921
Scn1a(2),9.224,7.369,10.145,14.242,17.262,11.535,8.144,7.166,13.625,6.604
Ntrk2(2),21.929,17.018,14.799,19.783,14.632,24.421,14.235,9.344,16.658,17.913
Cacna1c,4.585,3.637,3.948,4.135,3.403,5.381,4.193,3.162,3.455,3.695
Grin2b,3.273,2.57,2.101,2.922,1.826,3.338,2.121,1.416,2.973,2.005
Scn9a(2),0.319,0,0,0.453,0.434,0.376,0,0,0.346,0.469
Gria4(2),10.867,8.156,7.889,9.236,14.134,10.574,8.404,8.179,9.442,7.982
Cacna1e(2),1.805,1.783,2.045,1.968,1.405,1.807,0.973,0.993,0.857,1.769
Gria3,4.237,4.188,3.901,5.221,6.439,3.993,3.421,4.012,4.452,4.631
Gria1,8.284,7.942,7.557,12.001,3.976,9.472,7.653,4.16,7.971,5.381
Kcnj5,3.089,2.046,3.332,3.392,2.168,3.786,3.865,1.414,2.37,2.009
Cacna1b(2),11.071,8.716,8.246,9.594,7.189,11.62,6.028,4.481,9.307,9.074
Scn5a,1.301,1.017,0.714,1.401,0.449,1.183,1.065,0.292,0.823,0.714
Scn2a(2),3.286,2.119,2.257,2.024,1.902,3.441,1.327,1.072,2.576,2.09
Scn10a,0.037,0.069,0.087,0.076,0.082,0.095,0.052,0.019,0.078,0.045
Cacna1g(2),6.543,5.095,5.463,8.404,3.084,7.359,5.746,4.682,5.969,4.315
Cacna1e(3),5.37,4.002,3.313,4.803,2.665,5.623,3.296,1.953,3.827,4.092
Bdnf(4),0.869,0.509,0.996,1.032,0.256,0.742,0.498,0.531,0.994,0.473
Scn4a,0.284,0.278,0.359,0.45,0.761,0.31,0.319,0.27,0.366,0.273
Scn5a(2),0.256,0.477,0.587,0.283,0,0.564,0.044,0.023,0.204,0.15
Gabra1,51.019,44.3,57.609,81.522,40.853,64.921,68.263,31.766,58.006,39.518
Scn8a,6.854,14.666,5.416,12.347,4.823,14.935,7.014,16.684,9.686,17.44
Kcnj3,17.047,14.3,13.741,14.363,14.01,13.268,12.172,10.718,15.374,13.048
Slc6a2,107.9,69.941,91.704,36.411,112.57,114.5,23.398,63.848,53.323,135.26
Grin3a,6.952,5.676,7.301,12.557,3.65,10.628,9.783,4.286,8.015,4.499
Cnr1,20.261,16.981,19.996,26.469,12.709,24.705,25.548,10.61,19.746,14.64
Scn1b,13.732,15.763,5.03,20.68,17.788,14.959,16.298,24.682,22.477,15.117
Gria1(2),2.709,3.667,2.51,2.9,2.134,1.93,4.308,2.59,2.487,1.742
Scn3a(2),1.439,2.614,0,0.352,0,1.358,1.027,0,0.452,0.586
Scn11a,0.058,0.292,0.036,0.127,0.058,0.06,0.074,0.164,0.047,0.05
Gria1(3),25.283,17.779,22.725,32.705,8.823,28.727,26.915,12.876,23.545,17.879
Cacna1f,0.056,0.067,0.14,0.123,0.04,0.182,0.072,0.083,0.077,0.097
Cacna1a,20.791,19.816,17.613,21.663,15.697,22.824,16.737,16.719,16.604,20.469
Gria4,8.51,7.107,8.342,9.338,7.46,8.877,7.673,6.341,8.393,9.555
Scn8a,6.738,14.706,4.172,11.467,2.552,10.757,6.021,15.222,3.588,11.333
Grin2d,20.398,15.794,22.521,24.693,16.97,24.108,24.19,21.016,18.314,19.044
Gria3(2),15.301,13.087,13.918,14.433,12.282,14.914,12.198,11.602,13.738,15.481
Oprk1(2),6.66,4.97,7.604,10.281,2.151,10.462,10.278,1.525,6.869,4.902
Scn1b(3),46.553,42.795,49.498,55.558,64.101,38.178,44.1,59.033,43.837,39.382
Cacna1h,9.145,7.295,8.7,8.028,5.415,10.799,8.21,6.332,8.455,7.683
Scn2a,36.803,29.975,30.609,38.334,19.053,39.127,31.146,23.066,30.896,32.345
Cacna1g,5.489,5.213,6.24,7.896,3.97,4.876,6.283,5.464,6.08,3.692
Ntrk2(3),147.81,152.45,153.46,136.09,181.1,156.85,219.8,164.53,156.64,147.92
Scn1a,9.222,9.162,9.659,13.83,12.679,8.088,11.45,10.406,9.503,6.827
Grin1(3),69.943,68.01,76.358,81.029,63.692,83.424,70.981,80.088,69.821,70.764
Grin3b(2),2.065,1.265,1.45,1.576,3.875,1.441,1.822,1.964,2.286,0.965
Gabra2(2),2.268,1.251,1.638,2.844,2.93,2.934,3.725,1.724,1.455,2.674
Scn1b2(2),161.76,164.24,213.24,209.19,235.38,172.98,207.33,216.96,198.26,130.93
Oprm1(2),4.046,5.181,2.362,1.925,0.806,2.232,1.178,1.491,3.259,3.751
Cacna1c(3),0.077,0.194,0.23,0,0.132,0.127,0,0.035,0.09,0.092
Ntrk2,27.139,26.028,23.881,27.22,22.259,30.728,22.381,19.782,24.704,30.85
Cacna1d(2),2.126,2.263,2.038,2.1,1.995,2.966,1.943,2.01,2.317,2.214
Scn3a,21.272,16.356,16.245,14.875,11.825,19.753,10.994,11.08,16.905,19.832
Grin1(2),76.771,65.788,66.059,78.716,33.91,88.228,73.859,47.717,70.674,61.275
Grina,672.31,705.45,679.04,623.4,597.51,742.12,619.74,662.95,665.18,781.29
Cacna1e,2.448,1.981,1.506,2.003,1.318,3.052,1.953,0.814,2.17,2.482
Bdnf(2),1.853,2.128,2.553,1.996,0.663,2.5,2.385,0.468,1.922,1.481
Fos,18.402,24.653,23.038,20.615,8.027,38.444,20.836,11.756,20.823,20.296
Scn4b,23.772,27.874,25.388,25.109,51.926,20.291,25.521,28.701,30.256,17.344
Slc6a2(3),480.05,455.95,307.6,186.82,376.96,447.61,123.5,409.58,347.86,681.04
Ntf3(3),1.87,3.561,2.421,3.133,2.134,2.327,1.712,2.32,1.735,3.497
Bdnf(3),0.319,0.09,0.665,0.187,0.107,0.185,0.394,0.264,0.21,0.345
Scn3b,112.86,115.29,99.711,96.245,71.741,122.34,85.875,88.906,102.88,132.13
Grin2c,14.224,15.944,15.473,21.936,32.732,13.98,20.168,23.958,14.541,17.402
Gabrd,0.701,3.542,0.532,5.222,5.593,0.133,2.954,0.961,0.506,2.152
Cacna1b,16.935,15.764,14.475,15.639,10.655,19.408,14.115,14.079,14.26,16.737
Slc18a2,433.92,429.22,293.57,164.53,287.51,370.72,93.973,283.12,321.49,551.07
Cacnb1(2),16.456,5.099,16.969,4.469,12.471,5.143,14.017,10.049,17.537,4.26
Gabrg1,40.614,37.373,43.103,39.253,47.768,41.202,51.665,37.74,42.17,39.097
Grin1,1.235,0.812,0.909,1.605,0.513,1.371,1.596,1.346,1.213,0.922
Slc6a2(2),138.21,136.75,34.759,38.393,25.89,87.126,0,0.467,99.703,137.66
Galr3,2.691,2.51,2.517,4.446,0.727,2.933,4.041,2.08,2.638,1.456
Oprm1(3),7.273,7.676,7.08,6.196,5.515,9.023,2.57,4.8,7.699,10.471
Gabrq,70.623,67.728,51.095,42.456,43.156,77.924,28.63,32.975,54.192,87.697
Gria4(3),25.846,26.045,24.37,37.866,18.037,26.907,31.423,21.292,26.795,24.642
Cacna1c(2),0.644,0.894,0.831,1.084,0.721,1.026,0.817,0.371,1.333,1.015
Cacna1d(3),0.299,0.406,0.127,0.319,0.319,0.231,0.178,0.075,0.18,0.405
Cacnb1,47.24,51.505,42.702,48.718,33.28,60.334,38.611,41.827,40.352,56.132
Scn7a,2.351,2.38,2.114,1.96,0.316,2.647,1.945,1.219,2.559,1.498
Cacna1d,2.661,2.733,2.714,2.649,2.403,2.923,3.216,2.768,2.401,2.302
Gabra2,25.209,26.731,23.249,25.599,20.17,22.928,24.072,18.664,23.808,23.306
Scn9a,3.209,3.106,3.212,3.206,1.094,3.35,3.994,1.934,2.883,2.046
Ntf3(2),2.347,2.282,2.112,1.025,1.762,2.029,0.501,1.652,2.717,1.982
Gria2,12.726,12.997,12.74,15.615,7.156,14.375,13.387,11.682,12.968,11.332
Bdnf,0.703,0.777,1.034,0.571,0.166,1.164,0.549,0.325,0.801,1.12
Gria2(2),17.769,17.694,16.62,18.603,11.295,19.926,18.044,13.594,16.946,17.712
Bdnf(5),1.321,2.152,1.882,2.397,1.598,3.072,3.038,1.53,2.04,1.464

Answer 1

Here's a working sample using just base R. Using your goi:
str(goi)
# 'data.frame': 92 obs. of 11 variables:
# $ Gene : chr "Oprd1" "Grin2a" "Grin2d(2)" "Oprm1" ...
# $ Rat_1 : num 2.75 3.13 4.5 2 137.35 ...
# $ Rat_2 : num 1.39 2.64 5.53 1.8 97.16 ...
# $ Rat_3 : num 2.25 2.96 2.63 1.61 113.65 ...
# $ Rat_4 : num 3.36 5.17 4.68 1.71 141.93 ...
# $ Rat_5 : num 3.19 2.48 3.93 3.67 77.3 ...
# $ Rat_6 : num 2.43 3.54 6.05 3.21 133.02 ...
# $ Rat_7 : num 1.985 2.596 0.98 0.249 88.872 ...
# $ Rat_8 : num 1.75 1.535 0.077 1.248 75.586 ...
# $ Rat_9 : num 2.75 3.2 4.38 1.76 108.96 ...
# $ Rat_10: num 1.77 2.23 2.33 2.67 97.63 ...
control <- goi[,2:6]
stress <- goi[,7:11]
Now, instead of using for loop and processing each return as we calculate it, let's calculate everything, store the complete object for each test within the list, and preserve the opportunity to grab whatever we want from all tests afterwards.
results <- lapply(seq_len(nrow(goi)),
function(i) t.test(control[i,], stress[i,]))
length(results)
# [1] 92
Each element of results is the return value from a single call of t.test.
results[[1]]
# Welch Two Sample t-test
# data: control[i, ] and stress[i, ]
# t = 1.1034, df = 6.2218, p-value = 0.3107
# alternative hypothesis: true difference in means is not equal to 0
# 95 percent confidence interval:
# -0.5386851 1.4374851
# sample estimates:
# mean of x mean of y
# 2.5874 2.1380
You can access any component of the test results:
names(results[[1]])
# [1] "statistic" "parameter" "p.value" "conf.int" "estimate"
# [6] "null.value" "alternative" "method" "data.name"
head( sapply(results, `[[`, "p.value") )
# [1] 0.3107098 0.3083295 0.2626753 0.6245368 0.4406157 0.2800657
head( t(sapply(results, `[[`, "conf.int")) )
# [,1] [,2]
# [1,] -0.5386851 1.4374851
# [2,] -0.7513650 2.0681650
# [3,] -1.5018657 4.4862657
# [4,] -1.1880098 1.8504098
# [5,] -23.5402499 48.8678499
# [6,] -2.2762668 0.8250668
NB: one of R's many nuances is the fact that the *apply family will return a matrix that some might think is transposed from what it should be. Because f this, calls that return a matrix will benefit from being sandwiched in t(...). (This is a great opportunity to press the "I Believe" button and move on.)
You can combine all of these results into a single data.frame with something like:
namefunc <- function(x, nameroot) { dimnames(x) <- list(NULL, paste0(nameroot, seq_len(ncol(x)))) ; x ; }
(That was a small helper function to make the following slightly easier to read. It's a very naïve naming convention, used only to keep the columns unique for now.)
test_results <- cbind.data.frame(
statistic = sapply(results, `[[`, "statistic"),
p.value = sapply(results, `[[`, "p.value"),
parameter = sapply(results, `[[`, "parameter"),
namefunc( t(sapply(results, `[[`, "conf.int")), "conf" ),
namefunc( t(sapply(results, `[[`, "estimate")), "est" )
)
head(test_results)
# statistic p.value parameter conf1 conf2 est1 est2
# 1 1.1033554 0.3107098 6.221806 -0.5386851 1.4374851 2.5874 2.1380
# 2 1.0948456 0.3083295 7.312678 -0.7513650 2.0681650 3.2784 2.6200
# 3 1.2480711 0.2626753 5.480699 -1.5018657 4.4862657 4.2546 2.7624
# 4 0.5107431 0.6245368 7.337202 -1.1880098 1.8504098 2.1594 1.8282
# 5 0.8134064 0.4406157 7.633546 -23.5402499 48.8678499 113.4766 100.8128
# 6 -1.2161356 0.2800657 4.824393 -2.2762668 0.8250668 1.2900 2.0156
There is definitely room here to use packages from the tidyverse as RobertMc suggested. For that, I recommend dplyr and tidyr, though perhaps broom has utility here as well.