week 06

project_suggestions.md

@@ -0,0 +1,5 @@
+# Project suggestions
+
+
+* RNA-seq analysis with cell type deconvolution: Follow the approach in https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1008120 and analyze an RNA-seq dataset generated at the FGCZ, more infos from Hubert,  https://sta426hs2020.slack.com/archives/D01AE6J7EKD
+* 

week06-19oct2020/exercise/RSEM_counts/N1.stat/N1.cnt

@@ -0,0 +1,50 @@
+8625 2410661 0 2419286
+2355208 55453 1593695
+7276600 3
+0	8625
+1	816966
+2	549781
+3	347624
+4	237335
+5	158948
+6	83020
+7	62366
+8	42518
+9	27884
+10	22155
+11	17277
+12	13465
+13	9716
+14	8152
+15	3583
+16	3187
+17	2028
+18	1693
+19	941
+20	631
+21	382
+22	319
+23	154
+24	134
+25	106
+26	86
+27	37
+28	24
+29	3
+30	28
+31	3
+32	61
+34	4
+35	8
+36	1
+38	3
+39	20
+40	1
+41	4
+44	2
+45	1
+46	3
+51	1
+53	4
+64	2
+Inf	0

week06-19oct2020/exercise/RSEM_counts/N2.stat/N2.cnt

@@ -0,0 +1,52 @@
+8175 2411974 0 2420149
+2350559 61415 1600394
+7307445 3
+0	8175
+1	811580
+2	553246
+3	340463
+4	238469
+5	160161
+6	87691
+7	66420
+8	44923
+9	28260
+10	21095
+11	16988
+12	12137
+13	9229
+14	7423
+15	3853
+16	3215
+17	2042
+18	1699
+19	937
+20	676
+21	402
+22	337
+23	143
+24	164
+25	104
+26	107
+27	51
+28	21
+29	10
+30	22
+32	51
+33	4
+34	4
+35	4
+38	2
+39	20
+40	1
+41	6
+42	1
+45	1
+46	5
+47	1
+48	1
+50	1
+57	1
+64	2
+71	1
+Inf	0

week06-19oct2020/exercise/RSEM_counts/N3.stat/N3.cnt

@@ -0,0 +1,51 @@
+8100 2411735 0 2419835
+2344016 67719 1634599
+7382391 3
+0	8100
+1	777136
+2	566580
+3	357832
+4	240275
+5	161790
+6	88881
+7	63599
+8	43867
+9	28025
+10	22987
+11	17983
+12	10715
+13	9831
+14	6982
+15	4125
+16	3749
+17	2248
+18	1765
+19	1095
+20	656
+21	442
+22	339
+23	166
+24	187
+25	121
+26	104
+27	44
+28	23
+29	6
+30	33
+31	3
+32	64
+33	3
+35	11
+37	1
+38	3
+39	35
+40	1
+41	7
+45	1
+46	7
+48	4
+52	1
+53	3
+60	1
+64	4
+Inf	0

week06-19oct2020/exercise/RSEM_counts/T1.stat/T1.cnt

@@ -0,0 +1,54 @@
+8397 2412500 0 2420897
+2317309 95191 1599651
+7305161 3
+0	8397
+1	812849
+2	563605
+3	348478
+4	225290
+5	131070
+6	96701
+7	59981
+8	59209
+9	28054
+10	39675
+11	14158
+12	10035
+13	6407
+14	5108
+15	3213
+16	2541
+17	1841
+18	1443
+19	795
+20	578
+21	363
+22	303
+23	171
+24	139
+25	109
+26	100
+27	67
+28	36
+29	7
+30	35
+31	3
+32	51
+34	7
+35	6
+38	1
+39	16
+40	6
+41	9
+42	1
+45	9
+46	7
+47	2
+48	4
+49	1
+51	2
+53	5
+56	1
+57	1
+64	7
+Inf	0

week06-19oct2020/exercise/RSEM_counts/T2.stat/T2.cnt

@@ -0,0 +1,52 @@
+8495 2410446 0 2418941
+2310764 99682 1626085
+7354721 3
+0	8495
+1	784361
+2	564168
+3	361196
+4	242081
+5	139904
+6	89062
+7	59256
+8	56840
+9	26629
+10	36846
+11	13837
+12	11134
+13	6938
+14	5598
+15	3184
+16	2652
+17	1820
+18	1495
+19	827
+20	693
+21	597
+22	266
+23	236
+24	121
+25	257
+26	189
+27	79
+28	26
+29	8
+30	29
+31	5
+32	40
+33	2
+34	5
+35	9
+36	2
+39	8
+40	9
+41	4
+42	3
+45	4
+46	11
+48	2
+52	2
+53	3
+55	1
+64	7
+Inf	0

week06-19oct2020/exercise/RSEM_counts/T3.stat/T3.cnt

@@ -0,0 +1,54 @@
+8766 2412898 0 2421664
+2341964 70934 1607492
+7343248 3
+0	8766
+1	805406
+2	570918
+3	332612
+4	240601
+5	135160
+6	92354
+7	60493
+8	58647
+9	27500
+10	37988
+11	15284
+12	10816
+13	6989
+14	5630
+15	3471
+16	2844
+17	1892
+18	1434
+19	831
+20	629
+21	373
+22	297
+23	176
+24	117
+25	99
+26	83
+27	70
+28	25
+29	7
+30	31
+31	2
+32	38
+33	1
+34	2
+35	4
+36	3
+38	2
+39	29
+40	4
+41	6
+42	1
+45	3
+46	10
+48	5
+51	1
+53	3
+55	2
+60	2
+64	3
+Inf	0

week06-19oct2020/exercise/exercise06.Rmd

@@ -0,0 +1,47 @@
+---
+title: "Exercise 6"
+author: "Hubert Rehrauer"
+date: "19 10 2020"
+output: html_document
+---
+
+```{r setup, include=FALSE}
+knitr::opts_chunk$set(echo = TRUE)
+```
+
+
+## Comparison of Expression Estimates
+
+The data set consists for 3 Normal samples and 3 Treated samples with 2.5 Mio reads each. The reads are simulated to come from genes on the human chromosome 1.
+
+The expression values from featureCounts were created with:
+We use the library `Rsubread` to generate the counts
+```{r, eval=FALSE}
+library(Rsubread)
+countResult1 = featureCounts(..., strandSpecific=0,
+                                 GTF.featureType="exon", GTF.attrType="gene_id", useMetaFeatures=TRUE,
+                                 allowMultiOverlap=TRUE, countMultiMappingReads=FALSE)
+save(countResult1, file="countResult1.RData")
+countResult2 = featureCounts(..., strandSpecific=0,
+                                 GTF.featureType="exon", GTF.attrType="gene_id", useMetaFeatures=TRUE,
+                                 allowMultiOverlap=TRUE, countMultiMappingReads=TRUE, fraction=TRUE)
+save(countResult2, file="countResult2.RData")
+```
+
+The RSEM data were generated with
+```{sh, eval=FALSE}
+rsem-calculate-expression .... --calc-pme --calc-ci --strandedness none ...
+```
+For documentation see: http://deweylab.github.io/RSEM/rsem-calculate-expression.html
+
+
+
+## Exercises
+
+1. Depending on the featureCount mode, how many alignments are ignored during counting?  
+2. Compare the counts from featureCounts with the genelevel counts from RSEM. Which genes have different values?
+3. How does the sum of the counts in RSEM compare to the number of the aligned reads?  
+4. Compute pair-wise correlation of the samples based on the isoform counts and gene level counts
+
+
+