1 Introduction
In this analysis samples from mouse ameloblasts were sequenced using Oxford Nanopore direct RNA sequecing. In total, 2 replicates were sequenced. Poly(A) lengths distribution was analyzed for each sample, as well as expression.
1.1 Highlights
- 4 samples (2 TENT5A WT and 2 TENT5A KO) sequenced using Nanopore Direct RNA sequencing
- Significant poly(A) tail shortening identified for 14 transcripts, including: Amelx, Col1a1, or Sparc
- For Amelx, tail length is descreased for shorter isoforms
- Decrease of poly(A) length is accompanied by decreased expression
2 Nanopolish QC and samples summary
2.1 Samples and groups definition
In the curent analysis 6 samples were analyzed, arranged in 2 groups. The table with metadata for each sample is shown below.
| sample_name | group | replicate | comment | guppy | kit | flowcell | seq_date |
|---|---|---|---|---|---|---|---|
| TENT5A_WT_1 | TENT5A_WT | 1 | with yeast trf5delta | 4.4.1 | SQK-RNA002 | FLO-MIN106 | 27/11/2019 |
| TENT5A_WT_1 | TENT5A_WT | 1 | with yeast trf5delta, rerun of interrupted run | 4.4.1 | SQK-RNA002 | FLO-MIN106 | 27/11/2019 |
| TENT5A_WT_2 | TENT5A_WT | 2 | with yeast Mex67-AA Pab1deltaRRM | 3.6.0 | SQK-RNA002 | FLO-MIN106 | 15/05/2020 |
| TENT5A_KO_1 | TENT5A_KO | 1 | with yeast trf4 delta | 4.4.1 | SQK-RNA002 | FLO-MIN106 | 27/11/2019 |
| TENT5A_KO_1 | TENT5A_KO | 1 | with yeast trf4 delta, rerun of interrupted run | 4.4.1 | SQK-RNA002 | FLO-MIN106 | 27/11/2019 |
| TENT5A_KO_2 | TENT5A_KO | 2 | with yeast Mex67-AA Pab1deltaRRM | 3.6.0 | SQK-RNA002 | FLO-MIN106 | 15/05/2020 |
2.2 Nanopolish QC
For this analysis, a total of 5,137,279 reads were analyzed with Nanopolish and of these 3,682,890 (71.7%) were marked as satsifying the quality metric. The summary of QC is shown below for each analyzed sample.
2.3 Samples summary
Basic summary of obtained counts and mean poly(A) lengths for each sample is shown below. Only reads passing Nanopolish filter are included.
Although mean/median poly(A) lengths increase in TENT5A KO this is rather side-effect as further analyses (below) show transcripts with significant poly(A) length decrease upon TENT5A KO.
What is also visible, there is a strong batch effect, as for the second replicate tails are shorter. This was taken into account in statistical analysis.
Per trancript summary is available here
| sample_name | group | counts | polya_mean | polya_sd | polya_median | polya_gm_mean | polya_sem |
|---|---|---|---|---|---|---|---|
| TENT5A_KO_1 | TENT5A_KO | 902952 | 87.35166 | 58.52080 | 72 | 69.59917 | 0.0615854 |
| TENT5A_KO_2 | TENT5A_KO | 750814 | 78.33402 | 52.10089 | 65 | 62.79270 | 0.0601283 |
| TENT5A_WT_1 | TENT5A_WT | 401023 | 84.94418 | 54.41867 | 71 | 68.71915 | 0.0859337 |
| TENT5A_WT_2 | TENT5A_WT | 1628101 | 75.75083 | 47.80506 | 64 | 62.04084 | 0.0374656 |
| Note: | |||||||
| Polya_mean is the arithmetic mean of poly(A) lengths, whereas polya_gm_mean is a geometric mean of poly(A) lengths. polya_sd shows the standard deviation value. |
Group-level summary is shown below. Only reads passing Nanopolish filter are included.
Although mean/median poly(A) lengths increase in TENT5A KO this is rather side-effect as further analyses (below) show transcripts with significant poly(A) length decrease upon TENT5A KO.
Per trancript summary is available here
| group | counts | polya_mean | polya_sd | polya_median | polya_gm_mean | polya_sem |
|---|---|---|---|---|---|---|
| TENT5A_KO | 1653766 | 83.25763 | 55.87858 | 69 | 66.42208 | 0.0434519 |
| TENT5A_WT | 2029124 | 77.56774 | 49.31875 | 66 | 63.30713 | 0.0346224 |
| Note: | ||||||
| Polya_mean is the arithmetic mean of poly(A) lengths, whereas polya_gm_mean is a geometric mean of poly(A) lengths. polya_sd shows the standard deviation value. |
2.4 Samples correlation matrix
To assess inter- and intra-group variability, Spearman correlations were calculated for each pair of samples, taking into account transcript abundances (counts) and calculated poly(A) lengths. In the ideal case, samples from the same group should cluster together and have high correlation values
2.4.1 poly(A) lengths correlation
Spearman correlation matrix was calculated based on polya_gm_mean values, using transcripts with at least 20 reads per transcript in all samples.
2.4.2 counts correlation
Spearman correlation matrix was calculated using counts for each transcript in the samples.
2.5 Principal Component Analysis
To further assess samples similarity, Principal Component Analysis was calculated. In the ideal situation, all samples from the same group should cluster together on the plots.
2.5.1 counts
PCA was calculated using counts for each transcript in the samples.
2.5.2 poly(A) length
PCA was calculated based on polya_gm_mean values, using transcripts with at least 20 reads per transcript in all samples.
2.6 Transcript abundances
In total, 35609 transcripts were identified, with mean 103.4258193 and median 10 count for each transcript (cumulative for all samples). 6192 transcripts were represented with single read, 2988 with 2 and 17430 with less then 10 reads.
This suggests good coverage of the transcriptome.
Histogram of transcripts abundances, with log10 y-axis is shown below. Outliers (found with boxplot function) are excluded for clarity.
## Attempting web service request:
## http://uswest.ensembl.org:80/biomart/martservice?type=version&requestid=biomaRt&mart=ENSEMBL_MART_ENSEMBL
## V1
## 1 0.7
## BioMartServer running BioMart version: 0.7
## Mart virtual schema: default
## Mart host: http://uswest.ensembl.org:80/biomart/martservice
## Attempting web service request:
## http://uswest.ensembl.org:80/biomart/martservice?type=attributes&dataset=mmusculus_gene_ensembl&requestid=biomaRt&mart=ENSEMBL_MART_ENSEMBL&virtualSchema=default
## Attempting web service request:
## http://uswest.ensembl.org:80/biomart/martservice?type=filters&dataset=mmusculus_gene_ensembl&requestid=biomaRt&mart=ENSEMBL_MART_ENSEMBL&virtualSchema=default3 Poly(A) lengths distribution
3.1 Per sample
3.1.1 density plot
Plot below shows the scaled density of calculated polyaA lengths for each sample. Vertical lines represent median values for each sample.
There is a clear difference in poly(A) lengths between sequencing batches. This will be taken into account when calculating statistics.
3.1.2 violin plot
3.2 Per group
3.2.1 density plot
Plot below shows the scaled density of calculated polyA lengths for each group. Vertical lines represent median values for each group.
3.2.2 violin plot
3.3 Estimated polyA length to dwell time correlation
Correlation between polyA dwell time and Nanopolish estimated polyA length is 0.957. Dwell time is calculated based on raw Nanopore signal and its value represents the number of sampling steps required to read the whole poly(A) tail as it traverse through the pore. Below a scatter plot sampled from 10000 random reads is shown.
4 Statistics
4.1 per transcript
4.1.1 table
statistics were calucalted using glm (Generalized Linerar Model), using log2(polya_length) as the response value, and condition (TENT5A WT/KO) and replicate as predicators. Replicate was use din the glm formula to account for the batch effect.
If there were multiple isoforms for given transcript identified, they were treated individually (hence: per transcript analysis). However, isoform assignments may be innacurate
Excel file is available for download here
4.1.2 volcano
4.1.3 Gene set enrichment
Gene set enrichment analysis was done with g:profiler, Ensembl annotation version 102. Enrichment was checked for functional terms from databases:
Online version of enrichment analysis
4.1.3.1 summary plot
As expected, for genes with decreased poly(A) tails in TENT5A KO functions connected with extracellular space and biomineral tissue development are enriched. For genes with increased poly(A) house-keeping functions, connected with translation, are enriched, what can be treated as a background.
Mouse over individual points to see names of gene ontology Terms.
- Upper panel - enrichment analysis for genes with decreased poly(A) length in mutant
- Lower panel - enrichment analysis for genes with increased poly(A) length in mutant
4.2 per gene
4.2.1 table
If there were multiple isoforms for given transcript identified, there were collapsed. This may provide clearer results.
If there were multiple isoforms for given transcript identified, they were collapsed
Excel file is available for download here
4.2.2 volcano
4.2.3 MA-like plot
- TENT5A substrate defined as having significantly shortened poly(A) tail (adjusted p.value < 0.05, length difference >5, at least 10 reads in TENT5A WT) in TENT5A KO
- poly(A) length difference is a difference of medians for TENT5A KO/WT, for each transcript
## svg
## 24.2.4 Gene set enrichment
Gene set enrichment analysis was done with g:profiler. Enrichment was checked for functional terms from databases:
4.2.5 summary plot
- Upper panel - enrichment analysis for genes with decreased poly(A) length in mutant
- Lower panel - enrichment analysis for genes with increased poly(A) length in mutant
4.2.5.1 DAVID analysis
This is an alternative to the analysis shown above (with G:profiler).
## $plot
##
## $data
## # A tibble: 12 x 13
## Category Term Count X. PValue Genes List.Total Pop.Hits Pop.Total
## <fct> <chr> <int> <dbl> <dbl> <chr> <int> <int> <int>
## 1 GOTERM_C… GO:000… 10 76.9 5.99e-8 ENSMUSG0… 13 1753 19662
## 2 GOTERM_C… GO:003… 6 46.2 5.25e-7 ENSMUSG0… 13 294 19662
## 3 GOTERM_C… GO:000… 6 46.2 7.50e-7 ENSMUSG0… 13 316 19662
## 4 GOTERM_C… GO:000… 8 61.5 8.49e-6 ENSMUSG0… 13 1504 19662
## 5 GOTERM_C… GO:000… 5 38.5 4.18e-4 ENSMUSG0… 13 629 19662
## 6 GOTERM_C… GO:000… 3 23.1 1.13e-3 ENSMUSG0… 13 83 19662
## 7 GOTERM_C… GO:000… 2 15.4 1.22e-3 ENSMUSG0… 13 2 19662
## 8 GOTERM_C… GO:007… 7 53.8 2.79e-3 ENSMUSG0… 13 2674 19662
## 9 GOTERM_C… GO:001… 2 15.4 1.94e-2 ENSMUSG0… 13 32 19662
## 10 GOTERM_C… GO:000… 2 15.4 5.76e-2 ENSMUSG0… 13 97 19662
## 11 GOTERM_C… GO:003… 2 15.4 6.63e-2 ENSMUSG0… 13 112 19662
## 12 GOTERM_C… GO:007… 2 15.4 7.82e-2 ENSMUSG0… 13 133 19662
## # … with 4 more variables: Fold.Enrichment <dbl>, Bonferroni <dbl>,
## # Benjamini <dbl>, FDR <dbl>
##
## $data_compact
## # A tibble: 12 x 5
## Term Count PValue Fold.Enrichment Benjamini
## <chr> <int> <dbl> <dbl> <dbl>
## 1 GO:0005576~extracellular region 10 5.99e-8 8.63 3.95e-6
## 2 GO:0031012~extracellular matrix 6 5.25e-7 30.9 1.65e-5
## 3 GO:0005578~proteinaceous extrace… 6 7.50e-7 28.7 1.65e-5
## 4 GO:0005615~extracellular space 8 8.49e-6 8.05 1.40e-4
## 5 GO:0009986~cell surface 5 4.18e-4 12.0 5.52e-3
## 6 GO:0005581~collagen trimer 3 1.13e-3 54.7 1.15e-2
## 7 GO:0005584~collagen type I trimer 2 1.22e-3 1512. 1.15e-2
## 8 GO:0070062~extracellular exosome 7 2.79e-3 3.96 2.30e-2
## 9 GO:0016235~aggresome 2 1.94e-2 94.5 1.42e-1
## 10 GO:0005604~basement membrane 2 5.76e-2 31.2 3.80e-1
## 11 GO:0030141~secretory granule 2 6.63e-2 27.0 3.98e-1
## 12 GO:0072562~blood microparticle 2 7.82e-2 22.7 4.30e-15 Differential expression
Nanopore reads were mapped to the mouse genome (mm10) with Minimap2, counted with featureCounts (subread package) and analyzed with DESeq2 package. For the annotation, the Biomart was used.
5.1 mapping summary:
- fraction of unmapped reads is coming from yeast polyA+ RNA which were added to each library.
5.2 TENT5A WT vs KO
5.2.1 PCA plot
- Good separation between WT and KO samples
- batch effect (which was shown for poly(A) lengths) is visible on PC2
- Excel file is available for download here
5.2.2 Table
- only significantly changed genes (expression) shown below
5.2.3 Volcano
5.2.4 Enrichment analysis
Enrichment analysis shows similar ontology terms as the analysis done for genes identified in poly(A) lengths analysis
Mouse over points to see onotology names.
- upper panel - decreased expression
- lower panel - increased expression
5.2.5 Poly(A) lengths
5.2.5.1 Decreased expression in mutant
- top 50 transcripts (based on fold change)
5.2.5.1.1 Density plot
5.2.5.1.2 Boxplot
5.2.5.1.3 Table
5.2.5.2 Increased expression in mutant
- top 50 transcripts (based on fold change)
5.2.5.2.1 Density plot
5.2.5.2.2 Boxplot
5.2.5.2.3 Table
6 groups of transcripts
6.1 TENT5A substrates
6.1.1 boxplot
## svg
## 26.1.1.1 Violin plot
## svg
## 26.1.1.2 Beeswarm plot
## svg
## 26.1.1.3 poly(A) length comparison
Plots below show the difference in median poly(A) length between WT and KO. Horizontal bars inidicates the median length of poly(A) in WT (upper bar) or KO (bottom bar).
6.1.1.3.1 sorted by length_diff
## svg
## 26.1.2 Sorted by WT polyA length
## svg
## 26.1.3 Sorted by KO polyA length
## svg
## 26.2 Comparison of expression and poly(A) lengths
6.2.1 All genes with significantly changed expression
- Only genes with significantly changed expression (as revealed with DESEq2) are included
- TENT5A substrates marked in blue
## svg
## 26.2.2 All TENT5A substrates
- only TENT5A substrates shown
- transcripts with significantly changed expression (as revealed by DESeq2 analysis) are marked i nred
## svg
## 26.2.3 Median poly(A) vs expression
- all transcripts with normalized expression>10 shown.
- TENT5A substrates marked with orange
- TENT5A substrates have rather long poly(A) tails in WT
## svg
## 27 Selected transcripts - poly(A) ditributions
7.1 global (averaged for all samples)
## svg
## 27.2 Col1a1
7.2.1 Density plot
7.2.2 group density plot
## svg
## 27.2.3 Boxplot
7.3 mitochondrially-encoded - all protein coding
- their poly(A) lengths should be well-defined and constant between analyzed conditions
7.3.1 Density plot
7.3.2 group density plot
## svg
## 27.3.3 Boxplot
7.4 TENT5A substrates
- at least 5nt decrease in poly(A) length upon TENT5A KO, and significant (adjusted p.value <0.05) statistics
7.4.1 Density plot
7.4.2 group density plot
## svg
## 27.4.3 Boxplot
7.5 Amelx
7.5.1 Density plot
7.5.2 group density plot
## svg
## 27.5.3 Boxplot
7.6 Lyz2
7.6.1 Density plot
7.6.2 group density plot
## svg
## 27.6.3 Boxplot
7.7 Gapdh
7.7.1 Density plot
7.7.2 group density plot
## svg
## 27.7.3 Boxplot
7.8 Actb
7.8.1 Density plot
7.8.2 group density plot
## svg
## 27.8.3 Boxplot
7.9 ribosomal proteins transcripts
7.9.1 Density plot
7.9.2 group density plot
## svg
## 27.9.3 Boxplot
7.10 ApoE
7.10.1 Density plot
### group density plot
## svg
## 27.10.2 Boxplot
7.11 Etracellular region
- All extracellular region (GO:0005576) transcripts
- based on Gene Ontology (Ensembl 102)
- on plots all transcripts assigned to given functional term are shown, on other - all remaining transcripts
7.11.1 density plots
7.11.2 Boxplot
7.11.3 group boxplot
7.11.4 Transcripts list
- summary of transcripts in this functional term. All values are for the whole dataset (all samples, all groups)
7.12 Extracellular matrix transcripts
- All extracellular matrix (GO:0031012) transcripts
- based on GO
7.12.1 density plots
7.12.2 Boxplot
7.12.3 group boxplot
7.12.4 Transcripts list
- summary of transcripts in this functional term. All values are for the whole dataset (all samples, all groups)
8 Transcript properties - GC/UTRs etc
- Plots showing ditribution of UTRs lengths, coding sequences lengths or %GC content for TENT5A substrates (decreased poly(A) length) or transcripts with expression changed upon TENT5A KO in ameloblasts
8.1 poly(A) decreased
- poly(A) decreased in TENT5A KO (irrespective of polyA change, just with significant p.value)
8.1.1 Transcript length
8.1.1.1 density plot
8.1.1.2 violin plot
8.1.1.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: transcript_length by
## decreased (decreased poly(A), not decreased)
## Z = -1.2038, p-value = 0.2287
## alternative hypothesis: true mu is not equal to 08.1.2 3’UTR length
8.1.2.1 density plot
8.1.2.2 violin plot
8.1.2.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: utr_3_length by
## decreased (decreased poly(A), not decreased)
## Z = -1.6389, p-value = 0.1012
## alternative hypothesis: true mu is not equal to 08.1.3 5’UTR length
8.1.3.1 density plot
8.1.3.2 violin plot
8.1.3.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: utr_5_length by
## decreased (decreased poly(A), not decreased)
## Z = 0.056656, p-value = 0.9548
## alternative hypothesis: true mu is not equal to 08.1.4 GC content
8.1.4.1 density plot
8.1.4.2 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: percentage_gene_gc_content by
## decreased (decreased poly(A), not decreased)
## Z = 1.2588, p-value = 0.2081
## alternative hypothesis: true mu is not equal to 08.2 poly(A) increased
- poly(A) increased in TENT5A KO
8.2.1 Transcript length
8.2.1.1 density plot
8.2.1.2 violin plot
8.2.1.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: transcript_length by
## increased (increased poly(A), not increased)
## Z = -9.9977, p-value < 2.2e-16
## alternative hypothesis: true mu is not equal to 08.2.2 3’UTR length
8.2.2.1 density plot
8.2.2.2 violin plot
8.2.2.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: utr_3_length by
## increased (increased poly(A), not increased)
## Z = -6.5126, p-value = 7.385e-11
## alternative hypothesis: true mu is not equal to 08.2.3 5’UTR length
8.2.3.1 density plot
8.2.3.2 violin plot
8.2.3.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: utr_5_length by
## increased (increased poly(A), not increased)
## Z = -0.82894, p-value = 0.4071
## alternative hypothesis: true mu is not equal to 08.2.4 GC content
8.2.4.1 density plot
8.2.4.2 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: percentage_gene_gc_content by
## increased (increased poly(A), not increased)
## Z = 2.8611, p-value = 0.004221
## alternative hypothesis: true mu is not equal to 08.3 TENT5A substrates
- poly(A) decreased in TENT5A KO
- length_diff < -5
8.3.1 Transcript length
8.3.1.1 density plot
## svg
## 28.3.1.2 violin plot
## svg
## 28.3.1.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: transcript_length by substrate (other, TENT5A substrate)
## Z = 1.2038, p-value = 0.2287
## alternative hypothesis: true mu is not equal to 0##
## other TENT5A substrate
## 17134 148.3.2 CDS length
8.3.2.1 density plot
## svg
## 28.3.2.2 violin plot
## svg
## 28.3.2.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: cds_length by substrate (other, TENT5A substrate)
## Z = -0.31753, p-value = 0.7508
## alternative hypothesis: true mu is not equal to 0##
## other TENT5A substrate
## 17134 148.3.3 3’UTR length
8.3.3.1 density plot
## svg
## 28.3.3.2 violin plot
## svg
## 28.3.3.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: utr_3_length by substrate (other, TENT5A substrate)
## Z = 1.6389, p-value = 0.1012
## alternative hypothesis: true mu is not equal to 0##
## other TENT5A substrate
## 17134 148.3.4 5’UTR length
8.3.4.1 density plot
## svg
## 28.3.4.2 violin plot
## svg
## 28.3.4.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: utr_5_length by substrate (other, TENT5A substrate)
## Z = -0.056656, p-value = 0.9548
## alternative hypothesis: true mu is not equal to 0##
## other TENT5A substrate
## 17134 148.3.5 GC content
8.3.5.1 density plot
## svg
## 28.3.5.2 violin plot
## svg
## 28.3.5.3 statistics
##
## Asymptotic Wilcoxon-Mann-Whitney Test
##
## data: percentage_gene_gc_content by substrate (other, TENT5A substrate)
## Z = -1.2588, p-value = 0.2081
## alternative hypothesis: true mu is not equal to 0##
## other TENT5A substrate
## 17134 149 poly(A) sites usage
- selected transcripts, with significant changes in poly(A) lengths in TENT5AC (homozygous) mutant showed below.
- To get the polyadenylation site the bam files transcriptome-mapped nanopore reads where converted to bed (with bamToBed) and imported to R. The rightmost position of read is considered as the termination/polyadenylation site
- Peaks of coverage are then identified, to find possible alternative polyadenylation sites.
- Analysis of poly(A) lengths is performed by comparing WT/mutant and different poly(A) sites (if were identified)
9.1 Amelx
- found 4 possible polyadenylation sites, which differ with poly(A) lengths
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.2 Lyz2
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
## Col1a1
- similarly to osteoblasts, adenylation affected mostly on proximal poly(A) site
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.3 Col1a2
- again, two “proximal” peaks with decreased poly(A) lengths in KO
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
## Apoe
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.4 Bglap
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.5 Sparc
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.6 Ambn
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
## Gapdh
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.7 Dspp
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.8 Clu
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.9 Cpz
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
## Bgn
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.10 Pcolce
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
9.11 Bglap2
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
## Odam
## span increased to next odd value: 21
## span increased to next odd value: 21
## span increased to next odd value: 21
10 Reproducibility
Seqeuncing reads were basecalled with Guppy 4.4.1 and mapped to Gencode V26 reference transcriptome using MiniMap 2.17. Poly(A) lengths were estimated with Nanopolish 0.13.2. Visualizations and statistics were prepared using NanoTail package.
This report has been created using Rmarkdown and publicly available R packages for reproducibility. For clarity the R packages used, and their versions, is listed below.
R version 4.0.2 (2020-06-22)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 20.04.1 LTS
Matrix products: default
BLAS: /usr/local/software/R/4.0.2/lib/R/lib/libRblas.so
LAPACK: /usr/local/software/R/4.0.2/lib/R/lib/libRlapack.so
attached base packages:
[1] parallel stats4 stats graphics grDevices utils datasets methods base
loaded via a namespace (and not attached):
[1] utf8_1.2.1 tidyselect_1.1.0 RSQLite_2.2.6
[4] htmlwidgets_1.5.3 grid_4.0.2 BiocParallel_1.22.0
[7] modeest_2.4.0 munsell_0.5.0 codetools_0.2-18
[10] DT_0.18 withr_2.4.1 colorspace_2.0-0
[13] highr_0.8 knitr_1.32 rstudioapi_0.13
[16] assertive.base_0.0-9 ggsignif_0.6.1 rJava_0.9-13
[19] labeling_0.4.2 GenomeInfoDbData_1.2.3 bit64_4.0.5
[22] farver_2.1.0 fBasics_3042.89.1 TH.data_1.0-10
[25] vctrs_0.3.7 generics_0.1.0 xfun_0.22
[28] ggpmisc_0.3.9 BiocFileCache_1.12.1 R6_2.5.0
[31] ggbeeswarm_0.6.0 clue_0.3-58 assertive.sets_0.0-3
[34] locfit_1.5-9.4 bitops_1.0-6 cachem_1.0.4
[37] assertthat_0.2.1 promises_1.2.0.1 multcomp_1.4-16
[40] beeswarm_0.3.1 gtable_0.3.0 spatial_7.3-13
[43] sandwich_3.0-0 timeDate_3043.102 rlang_0.4.10
[46] genefilter_1.70.0 splines_4.0.2 rstatix_0.7.0
[49] rtracklayer_1.48.0 lazyeval_0.2.2 broom_0.7.6
[52] abind_1.4-5 yaml_2.2.1 modelr_0.1.8
[55] crosstalk_1.1.1 backports_1.2.1 httpuv_1.5.5
[58] RBGL_1.64.0 tools_4.0.2 ellipsis_0.3.1
[61] assertive.strings_0.0-3 jquerylib_0.1.3 RColorBrewer_1.1-2
[64] stabledist_0.7-1 splus2R_1.3-3 assertive.reflection_0.0-5
[67] Rcpp_1.0.6 progress_1.2.2 zlibbioc_1.34.0
[70] RCurl_1.98-1.3 prettyunits_1.1.1 ggpubr_0.4.0
[73] rpart_4.1-15 openssl_1.4.3 cowplot_1.1.1
[76] statip_0.2.3 zoo_1.8-9 haven_2.4.0
[79] ggrepel_0.9.1 cluster_2.1.1 fs_1.5.0
[82] assertive.models_0.0-2 assertive.data_0.0-3 timeSeries_3062.100
[85] openxlsx_4.2.3 reprex_2.0.0 mvtnorm_1.1-1
[88] mime_0.10 hms_1.0.0 evaluate_0.14
[91] xtable_1.8-4 XML_3.99-0.6 rio_0.5.26
[94] readxl_1.3.1 compiler_4.0.2 assertive.datetimes_0.0-3
[97] crayon_1.4.1 htmltools_0.5.1.1 later_1.1.0.1
[100] assertive_0.3-6 libcoin_1.0-8 geneplotter_1.66.0
[103] lubridate_1.7.10 DBI_1.1.1 assertive.files_0.0-2
[106] rmutil_1.1.5 dbplyr_2.1.1 MASS_7.3-53.1
[109] rappdirs_0.3.3 assertive.numbers_0.0-2 car_3.0-10
[112] cli_2.4.0 assertive.types_0.0-3 assertive.matrices_0.0-2
[115] assertive.data.uk_0.0-2 pkgconfig_2.0.3 GenomicAlignments_1.24.0
[118] coin_1.4-1 foreign_0.8-81 plotly_4.9.3
[121] xml2_1.3.2 assertive.data.us_0.0-2 annotate_1.66.0
[124] vipor_0.4.5 bslib_0.2.4 XVector_0.28.0
[127] AnnotationForge_1.30.1 rvest_1.0.0 digest_0.6.27
[130] assertive.code_0.0-3 Biostrings_2.56.0 rmarkdown_2.7
[133] cellranger_1.1.0 GSEABase_1.50.1 curl_4.3
[136] modeltools_0.2-23 shiny_1.6.0 Rsamtools_2.4.0
[139] lifecycle_1.0.0 jsonlite_1.7.2 carData_3.0-4
[142] viridisLite_0.4.0 askpass_1.1 fansi_0.4.2
[145] pillar_1.6.0 ggsci_2.9 lattice_0.20-41
[148] fastmap_1.1.0 httr_1.4.2 survival_3.2-10
[151] GO.db_3.11.4 glue_1.4.2 zip_2.1.1
[154] bit_4.0.4 Rgraphviz_2.32.0 assertive.properties_0.0-4
[157] stringi_1.5.3 sass_0.3.1 blob_1.2.1
[160] stable_1.1.4 memoise_2.0.0