## Sample Cat Treatment
## 1 Cat 2 CI 1
## 2 Cat 6 BI 1
## 3 Cat 11 CI 2
## 4 Cat 15 BI 2
## 5 Cat 20 CI 3
## 6 Cat 24 BI 3
## 7 Cat 29 CI 4
## 8 Cat 33 BI 4
## 9 Cat 38 CI 5
## 10 Cat 42 BI 5
## 11 Cat 47 CI 6
## 12 Cat 51 BI 6
## 13 Cat 56 CI 7
## 14 Cat 60 BI 7
## 15 Cat 2 CI 1
Two cat samples and each sample went under seven treatments. Based on the info we got from the training data we can assume There are seven cats which are treated with increased percentage of a compound in their diet. 7 different treatments (1-7, representing an increased percentage of a compound in their diet)
## Samples_ID Fwd_read_count Rev_read_count Total_count
## 1 Cat11 135705 135705 271410
## 2 Cat15 121938 121938 243876
## 3 Cat20 138085 138085 276170
## 4 Cat24 146494 146494 292988
## 5 Cat29 138384 138384 276768
## 6 Cat2 103819 103819 207638
## 7 Cat33 133808 133808 267616
## 8 Cat38 131974 131974 263948
## 9 Cat42 141054 141054 282108
## 10 Cat47 110455 110455 220910
## 11 Cat51 132224 132224 264448
## 12 Cat56 104119 104119 208238
## 13 Cat60 90326 90326 180652
## 14 Cat6 120924 120924 241848
summary(assessment.raw.fastq.counts$Total_count)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 180700 226100 264200 249900 275000 293000
- Does the numbers reflect similar HiSeq / MiSeq runs? MiSeq
Summary of FastQC reports are here
- Does the 52% GC content reflect microbial samples? No completely random
- Does the graphs show what is expected.
- Differences in forward and reverse read qualities.
- Is there anything to be worried about the per tile quality heatmap
plot?
- A very mixed per base sequence content distribution.
- High duplication levels.
- High levels of overrepresentation of sequences.
- Read lengths are 300bp. Is this normal to expect form MiSeq runs? How should we set our filtering (done after merging) based on this.
- Which variable region does our reads hit? Does it make sense in terms of the sequence technology/platform used?
## Samples_ID read_count
## 1 Cat11.merged.filtered_3 116366
## 2 Cat15.merged.filtered_3 110756
## 3 Cat20.merged.filtered_3 115317
## 4 Cat24.merged.filtered_3 128747
## 5 Cat29.merged.filtered_3 120693
## 6 Cat2.merged.filtered_3 88062
## 7 Cat33.merged.filtered_3 117888
## 8 Cat38.merged.filtered_3 114377
## 9 Cat42.merged.filtered_3 123856
## 10 Cat47.merged.filtered_3 94702
## 11 Cat51.merged.filtered_3 115170
## 12 Cat56.merged.filtered_3 90802
## 13 Cat60.merged.filtered_3 76637
## 14 Cat6.merged.filtered_3 107248
summary(filtered_fastqs.counts$read_count)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 76640 97840 114800 108600 117500 128700
On average ~13% of reads are dropped after filtering (we can recheck this)
(124950-108600)/124950
## [1] 0.1308523
Summary of FastQC reports on filtered/trimmed reads are here
- We still have a 52% GC content reflect microbial samples.
- Does the graphs show what is expected.
- Now we have high quality reads only. There is also nono low
hanging tails. Forward and reverse reaeds are now merged.
- We still have a verry mixed per base sequence content distribution.
- High duplication levels remains.
- High levels of overrepresentation remains.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent