Data-projects-with-R-and-GitHub
Background of the dataset
This data was acquired from two groups of healthy pregnant woman one with normal weight (NW, BMI 18.5-24.9) and one slightly overweight (OW, BMI 25-29.9). the researchers looked at the mothers immune cells to find out how the difference in BMI affect those cells during pregnancy and how those changes can link back to the diet, microbiome or baby microbiome.
“Rep” means: “biological replication” = Repeating an experiment using different biological specimens under the same conditions to account for natural variability
Source of the Data: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE273780
The columns of the data set each correspond to one sample, so G92 could be pregnant woman at week 36 normal weight rep 2 or similar. For later we are only interested in some of these observations.
To find out which conditions correspond to which columns go to the
source of the data.
If you are on the site which I linked under “Source of Data” you have to
scroll down until you see Samples (59) there you can press more and you
will see all the sample. Choose the sample you are interested in. You
will be redirected to a new page, where you just have to look for the
description. The description will tell you the name of the column by
that you can match column name and sample type
## X Row.names G58 G52 G25 G84 G2 G60 G41 G36 G73 G79
## 1 1 ENSG00000000003 23 0 23 7 14 18 15 3 0 12
## 2 2 ENSG00000000005 0 0 0 0 0 0 0 0 0 0
## 3 3 ENSG00000000419 469 348 494 334 567 616 342 383 363 409
## 4 4 ENSG00000000457 196 198 403 202 382 166 240 168 124 244
## 5 5 ENSG00000000460 40 45 82 38 70 36 40 26 33 52
## 6 6 ENSG00000000938 9567 14297 8058 13908 10128 6291 14438 10547 10985 15057
## 7 7 ENSG00000000971 20 3 5 4 31 22 1 22 14 12
## 8 8 ENSG00000001036 392 714 297 469 408 354 545 370 442 589
## 9 9 ENSG00000001084 330 337 490 320 489 374 421 402 279 416
## 10 10 ENSG00000001167 466 491 724 433 660 647 641 556 502 443
## G74 G90 G31 G46 G83 G89 G67 G28 G55 G26 G63 G69
## 1 5 4 11 19 8 3 8 16 24 10 27 16
## 2 0 0 0 0 0 0 0 0 0 0 0 0
## 3 402 317 382 387 443 424 444 632 813 373 1247 375
## 4 245 137 205 193 313 236 297 270 559 295 455 179
## 5 73 21 29 29 66 68 68 54 119 60 107 40
## 6 17799 13046 11323 11466 12620 23353 15412 13253 15210 12301 19379 17768
## 7 8 8 29 10 11 21 24 13 34 19 33 14
## 8 559 555 476 631 489 790 557 576 616 386 1174 642
## 9 500 240 334 357 426 370 464 490 867 460 852 334
## 10 540 394 421 494 596 454 503 495 1237 629 808 357
## G87 G48 G35 G42 G77 G38 G45 G32 G64 G80 G88 G11 G66
## 1 3 16 23 4 5 16 6 14 4 7 3 34 15
## 2 0 0 0 0 0 0 0 0 0 0 0 0 0
## 3 290 387 443 345 470 555 399 480 557 418 298 1024 386
## 4 191 187 165 213 340 356 326 388 261 297 165 633 190
## 5 60 28 23 40 71 78 90 84 65 37 33 132 22
## 6 19190 11581 7793 13080 29079 14459 9630 14931 12419 17088 18538 52130 9577
## 7 11 15 9 21 11 5 6 31 2 15 4 43 25
## 8 617 630 342 563 1107 638 494 558 498 760 668 2263 347
## 9 340 385 380 308 538 429 536 650 463 321 298 1195 351
## 10 441 548 550 563 648 678 860 1111 646 466 448 1110 496
## G54 G75 G30 G40 G78 G72 G59 G61 G9 G85 G81 G65 G12
## 1 15 1 3 24 8 7 16 27 13 0 18 11 6
## 2 0 0 0 0 0 0 0 0 0 0 0 0 0
## 3 530 359 450 422 362 417 468 589 631 319 426 473 372
## 4 306 265 278 385 290 195 253 309 400 133 253 306 192
## 5 65 84 61 69 75 35 82 53 99 24 49 96 54
## 6 15986 15436 13094 14061 9874 12515 11256 12136 8428 15919 14224 13083 8108
## 7 24 14 48 3 2 16 26 18 33 6 14 10 11
## 8 563 565 436 486 383 410 505 504 474 623 556 573 390
## 9 551 451 309 571 444 365 418 451 577 261 356 561 331
## 10 668 568 435 712 712 555 617 539 653 485 463 668 463
## G57 G76 G44 G43 G34 G49 G50 G27 G68 G86 G62 hgnc_symbol
## 1 1 4 21 18 2 15 0 7 4 12 13 TSPAN6
## 2 0 0 0 0 0 0 0 0 0 0 0 TNMD
## 3 374 390 518 426 389 483 375 525 435 460 514 DPM1
## 4 271 251 327 309 153 317 246 281 329 194 272 SCYL3
## 5 37 38 110 92 26 47 64 58 47 48 53 C1orf112
## 6 12489 16111 9555 11428 12809 10459 19620 13108 15107 7133 12090 FGR
## 7 45 10 13 19 15 7 5 17 26 9 31 CFH
## 8 527 535 398 456 646 316 746 524 641 383 464 FUCA2
## 9 400 338 496 533 408 367 535 394 298 428 406 GCLC
## 10 545 387 736 714 613 407 616 422 427 515 396 NFYA
## description
## 1 tetraspanin 6 [Source:HGNC Symbol;Acc:HGNC:11858]
## 2 tenomodulin [Source:HGNC Symbol;Acc:HGNC:17757]
## 3 dolichyl-phosphate mannosyltransferase subunit 1, catalytic [Source:HGNC Symbol;Acc:HGNC:3005]
## 4 SCY1 like pseudokinase 3 [Source:HGNC Symbol;Acc:HGNC:19285]
## 5 chromosome 1 open reading frame 112 [Source:HGNC Symbol;Acc:HGNC:25565]
## 6 FGR proto-oncogene, Src family tyrosine kinase [Source:HGNC Symbol;Acc:HGNC:3697]
## 7 complement factor H [Source:HGNC Symbol;Acc:HGNC:4883]
## 8 alpha-L-fucosidase 2 [Source:HGNC Symbol;Acc:HGNC:4008]
## 9 glutamate-cysteine ligase catalytic subunit [Source:HGNC Symbol;Acc:HGNC:4311]
## 10 nuclear transcription factor Y subunit alpha [Source:HGNC Symbol;Acc:HGNC:7804]
## chromosome_name strand external_gene_name
## 1 X -1 TSPAN6
## 2 X 1 TNMD
## 3 20 -1 DPM1
## 4 1 -1 SCYL3
## 5 1 1 C1orf112
## 6 1 -1 FGR
## 7 1 1 CFH
## 8 6 -1 FUCA2
## 9 6 -1 GCLC
## 10 6 1 NFYA
Wrangling goals
- Import the data
- Figure out what you are looking
- Filter out the count data of 2 replicates each of normal weight and overweight woman at week12 and at week 36 (G columns with number are the counts, and rows are the specific genes) so that you have the rows left for week12(Normal‑1, Normal‑2, Overweight‑1, Overweight‑2) and for week36 the same ones
- Keep the column, external gene names
- rename the columns with more readable names
- Normalize the data from counts to counts per million (add up all the counts per sample and divide each sample by this, then multiply by 1.000.000)
- log2 transformation of the counts per million to make the data more symmetric
- Drop any rows with blank names and count values between 0 and 2
- Calculate the z-Value of each count in new columns by subtracting the normalized count value by the mean of the column and divide by the columns standard deviation
- Calculate the variance of the log2 transformed counts by rows and put it in a new column.
With these steps you should end-up with the z-values and the varaince which are used in the heatmap
Visualization Goals
Produce a clustered heatmap of the top50 genes
- Rank all genes descendingly by variance
- keep only top 50 highest-variance genes for the heat map
- Rows should be the external gene names; columns the samples
- produce a heat map looking something like blow.
You can use this resource to get more information on how to create heatmaps: https://r-graph-gallery.com/heatmap
If you get a simple heatmap with just the annotations, that would already be amazing.
Further if you have a lot of time you can create an interactive heatmap. In which you can hover over the tiles and get information on the z-value, the gene name and the sample description. An example is also shown on the resource for heatmaps form above.
Why do such a data analysis?
At the end you could theoretically tell the difference between both groups and see if certain genes are enriched. For example inflammatory genes could be enriched but it is also possible that no difference or only slight differences will be reveal. These can be starting points for further experiments.
I hope it is going to work out for you!
Good Luck!!
