reemasingh / transcriptomics-assembly

De-novo Transcriptomics Assembly workflow for four Dictyostelium species (e.g.- Dictyostelium discoideum, Polysphondylium pallidum, Dictyostelium Lacteum and Dictyostelium Fasciculatum). This is the standard assembly workflow that should ideally work on any organism.Before normalizing first concatenate all RNAseq data across all samples into a single set of inputs to generate a single reference transcriptomics assembly. Combine all left reads in one file and all right reads in another file. In order to reduce the number, raw reads were normalized using in silico digital normalization implemented in trinity at 50X coverage. The reads were assembled with Trinity using kmer parameter of 25.

Transcriptomics-Assembly

De-novo Transcriptomics Assembly workflow for four Dictyostelium species (e.g.- Dictyostelium discoideum, Polysphondylium pallidum, Dictyostelium lacteum and Dictyostelium fasciculatum). This is the standard assembly workflow that should ideally work on any organism.Before normalizing first concatenate all RNAseq data across all samples into a single set of inputs to generate a single reference transcriptomics assembly. Combine all left reads in one file and all right reads in another file. In order to reduce the number, raw reads were normalized using in silico digital normalization implemented in trinity at 50X coverage. The reads were assembled with Trinity using kmer parameter of 25.

Alterative Link

https://github.com/bartongroup/rs-Transcriptomics-Assembly

Read Normalization

trinity/util/normalize_by_kmer_coverage.pl --seqType fq --JM 10G --max_cov 75 --left All_Left_Reads.fq --right All_Right_Reads.fq --pairs_together --PARALLEL_STATS --JELLY_CPU 6

Read Assembly using Trinity

alignReads.pl --left reads.ALL.left.fq.normalized_K25_C50_pctSD200.fq --right reads.ALL.right.fq.normalized_K25_C50_pctSD200.fq --seqType fq --target NewAssembly_35631.fasta.clean --aligner bowtie --retain_intermediate_files

Remove SpikeIns from Trinity Assembly

python remove_fasta.py

Assembly Statistics

trinity/util/alignReads.pl --left All_Left.fq --right All_Right.fq --seqType fq --target Trinity.fasta --aligner bowtie -- -p 6
samtools view -h -o bowtie_out.nameSorted.sam bowtie_out.nameSorted.bam
SAM_nameSorted_to_uniq_count_stats.pl bowtie_out/bowtie_out.nameSorted.sam >AlignmentStatistics

PASA Assembly

PASA/misc_utilities/accession_extractor.pl < trinity_out_dir/Trinity.fasta >tdn.accs
PASA/scripts/Launch_PASA_pipeline.pl -c alignAssembly.config -C -R -g PN500.fa -t NewAssembly_35631.fasta.clean --TDN tdn.accs --TRANSDECODER --ALT_SPLICE --ALIGNERS blat,gmap

PASA updation with the existing Annotation

In case the existing genome annotation in Augustus file format

grep "AUGUSTUS" PN500_augustus_prediction.gff | awk -F "\t" '$3 ~/gene|transcript|exon|CDS/' >P_Pal.gff 
python Format_Gff.py >P_Pal_1.gff
python Format_Gff2.py >P_Pal_2.gff
PASA/scripts/Load_Current_Gene_Annotations.dbi -c alignAssembly.config -g PN500.fa -P P_Pal_2.gff
PASA/scripts/Launch_PASA_pipeline.pl -c annotCompare.config  -A -g PN500.fa -t NewAssembly_35631.fasta.clean

Quality Control Measurement

Annotation

blat PN500.fa ../NewAssembly_35631.fasta.clean -t=dna -q=dna -out=psl Trinity-Genome.psl
perl blat2gbrowse.pl Trinity-Genome.psl Reema1.gff
less Reema1.gff | grep "HSP" >NewAssembly.gtf
perl GTF1.pl NewAssembly.gtf >NewAssembly_Final.gtf

Blobplot

blastn -task megablast -query Trinity.fasta -db nt -evalue 1e-5 -num_threads 6 -max_target_seqs 1 -outfmt '"6 qseqid staxids"' -out trinity.nt.1e-5.megablast
blobology/gc_cov_annotate.pl --blasttaxid trinity.nt.1e-5.megablast --assembly Trinity.fasta --bam bowtie_out.nameSorted.bam --out blobplot.txt --taxdump ./ --taxlist species order phylum superkingdom
blobology/makeblobplot.R blobplot.txt 0.01 taxlevel_order

samtools idxstats bowtie_out.coordSorted.bam >Read_Count_PerContig
grep 'Dictyosteliida\|Not annotated' blobplot.txt >blobplot_filter.txt
less blobplot_filter.txt | grep "Not annotated" | cut -f1 |sort -u >blobplot_filter_Not_Annot
un <- read.table(file="blobplot_filter_Not_Annot")
gc <- read.table(file="trinity_gc",header=TRUE)
cont <- read.table(file="Read_Count_PerContig",header=TRUE)
A1 <- gc[which(gc$Name %in% un$V1),]
Merge <- cbind(A1,cont[match(A1$Name,cont$Contig_Name),])
write.table(Merge,file="blobplot_filter_Not_Annot_GC_count.txt",sep="\t",row.names=FALSE,quote=FALSE)

awk '$3 >=55 && $6<=10' blobplot_filter_Not_Annot_GC_count.txt >blobplot_filter_Not_Annot_GC_count_filter.txt
cut -f1 blobplot_filter_Not_Annot_GC_count_filter.txt >blobplot_filter_Not_Annot_GC_count_filter
cut -f1 blobplot_filter.txt >blobplot_filter
full <- read.table(file="blobplot_filter")
filter <- read.table(file="blobplot_filter_Not_Annot_GC_count_filter")
A1 <- data.frame(full[-which(full$V1 %in% filter$V1),])
write.table(A1,file="Trinity_blobplot_filter",row.names=FALSE,col.names=FALSE,quote=FALSE)

perl -ne 'if(/^>(\S+)/){$c=$i{$1}}$c?print:chomp;$i{$_}=1 if @ARGV' Trinity_blobplot_filter Trinity.fasta >Trinity_blobplot_filter.fasta

CEGMA

CEGMA_v2.5/bin/cegma -g Trinity.fasta >trinity.cegma
CEGMA_v2.5/bin/cegma -g DF_genome.fa >genome.cegma
CEGMA_v2.5/bin/cegma -g pasa_Fasiculatum.assemblies.fasta >pasa1.cegma
CEGMA_v2.5/bin/cegma -g PASA2.fasta >pasa2.cegma

library(ggplot2)
a <- read.table(file="Cegma-Dicty1.txt",header=TRUE)
b <- read.table(file="cegma-Pal.txt",header=TRUE)
c <- read.table(file="cegma-Fasc.txt",header=TRUE)
d <- read.table(file="cegma-Lactum.txt",header=TRUE)

### Cegma-Fasc.txt contains the complete and partial CEG in all four files

require(ggplot2)
theme_set(theme_bw(14))
p <- ggplot(data=a,aes(x=Name,y=Score,fill=Type))+geom_bar(stat="identity")+xlab("")+ggtitle("Dictyostelium discoideum") + ylab ("Number of CEGs")+theme(text = element_text(size=20))+scale_fill_manual(values = c("skyblue3","skyblue2"))
p1 <- ggplot(data=b,aes(x=Name,y=Score,fill=Type))+geom_bar(stat="identity")+xlab("")+ggtitle("Polysphondylium pallidum") +ylab ("Number of CEGs")+theme(text = element_text(size=20))+ scale_fill_manual(values = c("burlywood3","burlywood2"))
p2 <- ggplot(data=c,aes(x=Name,y=Score,fill=Type))+geom_bar(stat="identity")+xlab("")+ggtitle("Dictyostelium fasciculatum") +ylab ("Number of CEGs")+theme(text = element_text(size=20))+ scale_fill_manual(values = c("indianred3","indianred2"))
p3 <- ggplot(data=d,aes(x=Name,y=Score,fill=Type))+geom_bar(stat="identity")+xlab("")+ggtitle("Dictyostelium lacteum") +ylab ("Number of CEGs")+theme(text = element_text(size=20))+scale_fill_manual(values = c("palegreen3","palegreen2"))

source("multiplot.R")
pdf("Cegma_staggerd.pdf",width=15,height=15)
multiplot(p,p2,p1,p3)
dev.off()

Transrate

transrate --assembly NewAssembly_35631.fasta.clean --left reads.ALL.left.fq.normalized_K25_C50_pctSD200.fq --right reads.ALL.right.fq.normalized_K25_C50_pctSD200.fq --reference PN500_augustus_prediction_test.aa --outfile Reference_Based
transrate --assembly DNA.fas --left reads.ALL.left.fq.normalized_K25_C50_pctSD200.fq --right reads.ALL.right.fq.normalized_K25_C50_pctSD200.fq --reference PN500_augustus_prediction_test.aa --outfile Reference_Based
transrate --assembly pasa_Pallidum_Gernot.assemblies.fasta --left reads.ALL.left.fq.normalized_K25_C50_pctSD200.fq --right reads.ALL.right.fq.normalized_K25_C50_pctSD200.fq --reference PN500_augustus_prediction_test.aa --outfile Reference_Based
transrate --assembly PASA2.fasta --left reads.ALL.left.fq.normalized_K25_C50_pctSD200.fq --right reads.ALL.right.fq.normalized_K25_C50_pctSD200.fq --reference PN500_augustus_prediction_test.aa --outfile Reference_Based

Plots

Assembly Depth Plots

dicty <- read.table(file="NewAssembly_31259_LengthnCount",sep="\t")
pal <- read.table(file="Pallidum_Length_Count",sep="\t")
fas <- read.table(file="Dfas_Length_Count",sep="\t")
lac <- read.table(file="Dlac_Length_Count",sep="\t")

t_frame <- data.frame(Species="D.discoideum",Count=dicty$V3)
p1_frame <- data.frame(Species="P.pallidum",Count=pal$V3)
p2_frame <- data.frame(Species="D.fasciculatum",Count=fas$V3)
d_frame <- data.frame(Species="D.lacteum",Count=lac$V3)
all_frame <- rbind(t_frame,p1_frame,p2_frame,d_frame)

library(ggplot2)
pdf("All_depth.pdf")
ggplot(all_frame, aes(x=Species, y=Count,fill=Species)) +geom_boxplot()+ scale_y_continuous(trans=log10_trans())+theme_bw()+theme(axis.line = element_line(colour = "black"),panel.grid.major = element_blank(),panel.grid.minor = element_blank(),panel.border = element_blank(),panel.background = element_blank())
dev.off()

CEGMA Plots

a <- read.table(file="Cegma-Dicty1.txt",header=TRUE)
b <- read.table(file="cegma-Pal.txt",header=TRUE)
c <- read.table(file="cegma-Fasc.txt",header=TRUE)
d <- read.table(file="cegma-Lactum.txt",header=TRUE)

require(ggplot2)
theme_set(theme_bw(14))
p <- ggplot(data=a,aes(x=Name,y=Score,fill=Type))+geom_bar(stat="identity")+xlab("")+ggtitle("Dictyostelium discoideum")+ylab("Number of CEGs")+theme(text = element_text(size=20))+scale_fill_manual(values = c("skyblue3","skyblue2"))
p1 <- ggplot(data=b,aes(x=Name,y=Score,fill=Type))+geom_bar(stat="identity")+xlab("")+ggtitle("Polysphondylium pallidum")+ylab("Number of CEGs")+theme(text = element_text(size=20))+ scale_fill_manual(values = c("burlywood3","burlywood2"))
p2 <- ggplot(data=c,aes(x=Name,y=Score,fill=Type))+geom_bar(stat="identity")+xlab("")+ggtitle("Dictyostelium fasciculatum")+ylab("Number of CEGs")+theme(text = element_text(size=20))+ scale_fill_manual(values = c("indianred3","indianred2"))
p3 <- ggplot(data=d,aes(x=Name,y=Score,fill=Type))+geom_bar(stat="identity")+xlab("")+ggtitle("Dictyostelium lacteum")+ylab("Number of CEGs")+theme(text = element_text(size=20))+scale_fill_manual(values = c("palegreen3","palegreen2"))

source("multiplot.R")
pdf("Cegma_staggerd.pdf",width=15,height=15)
multiplot(p,p2,p1,p3)
dev.off()

Transrate

Assembly Score

pal <- read.table(file="QA_Ref_Coveragre.txt",sep="\t",header=TRUE)
dp1_frame <- data.frame(name="Ddis_PASA1",n=pal$X,N=pal$DPasa1)
pp1_frame <- data.frame(name="Ppal_PASA1",n=pal$X,N=pal$PPasa1)
Fp1_frame <- data.frame(name="DFas_PASA1",n=pal$X,N=pal$FPasa1)
lp1_frame <- data.frame(name="DLac_PASA1",n=pal$X,N=pal$LPasa1)
dp2_frame <- data.frame(name="Ddis_PASA2",n=pal$X,N=pal$DPasa2)
lp2_frame <- data.frame(name="DLac_PASA2",n=pal$X,N=pal$LPasa2)
pp2_frame <- data.frame(name="Ppal_PASA2",n=pal$X,N=pal$PPasa2)
Fp2_frame <- data.frame(name="DFas_PASA2",n=pal$X,N=pal$FPasa2)

all_frame <- rbind(dp1_frame,pp1_frame,Fp1_frame,lp1_frame,dp2_frame,pp2_frame,Fp2_frame,lp2_frame)
library(ggplot2)
library(scales)
p <- ggplot(all_frame,aes(x=n,y=N,group=name,shape=name))+geom_point(aes(colour=name),size=4)+geom_line(aes(colour=name, group=name))+scale_color_manual(values=c("#FF3333","#0000CC","#00FF00","#FF00CC","#FF3333","#0000CC","#00FF00","#FF00CC"))+scale_shape_manual(values = c(1,1,1,1,16,16,16,16))+expand_limits(y=0)+xlab("Reference Coverage") + ylab("Proportion of reference transcripts")+ theme_bw() +theme(axis.line = element_line(colour = "black"),panel.grid.major = element_blank(),panel.grid.minor = element_blank(),panel.border = element_blank(),panel.background = element_blank(),text = element_text (size=20), legend.position ="top")+geom_hline(yintercept=c(0.22,0.35),linetype="dotted",color=c("#006633","#66CCFF"))

pal <- read.table(file="QA_Score.txt",sep="\t",header=TRUE)
dp1_frame <- data.frame(name="Ddis_PASA1",n=pal$X,N=pal$DPasa1)
pp1_frame <- data.frame(name="Ppal_PASA1",n=pal$X,N=pal$PPasa1)
Fp1_frame <- data.frame(name="DFas_PASA1",n=pal$X,N=pal$FPasa1)
lp1_frame <- data.frame(name="DLac_PASA1",n=pal$X,N=pal$LPasa1)
dp2_frame <- data.frame(name="Ddis_PASA2",n=pal$X,N=pal$DPasa2)
lp2_frame <- data.frame(name="DLac_PASA2",n=pal$X,N=pal$LPasa2)
pp2_frame <- data.frame(name="Ppal_PASA2",n=pal$X,N=pal$PPasa2)
Fp2_frame <- data.frame(name="DFas_PASA2",n=pal$X,N=pal$FPasa2)

all_frame <- rbind(dp1_frame,pp1_frame,Fp1_frame,lp1_frame,dp2_frame,pp2_frame,Fp2_frame,lp2_frame)
p1 <- ggplot(all_frame,aes(x=n,y=N,group=name,shape=name))+geom_point(aes(colour=name),size=4)+geom_line(aes(colour=name, group=name))+scale_color_manual(values= c("#FF3333", "#0000CC", "#00FF00", "#FF00CC", "#FF3333", "#0000CC", "#00FF00", "#FF00CC")) +scale_shape_manual(values = c(1, 1, 1, 1, 16, 16,16,16)) +expand_limits(y=0)+ xlab(" ") +ylab("Score")+theme_bw() +theme ( axis.line = element_line(colour = "black"),panel.grid.major = element_blank(),panel.grid.minor = element_blank(),panel.border = element_blank(), panel.background = element_blank(),text = element_text (size=20)) + geom_hline (yintercept = c(0.22,0.35) ,linetype ="dotted",color=c("#006633","#66CCFF"))

source("../multiplot.R")
pdf("Assembly_Score.pdf",width=15,height=15)
multiplot(p,p1)
dev.off()

Contig Score

Dictyostelium Discoideum

trinity <- read.table(file="Ddis/Reference_Based_NewAssembly_31259.fasta_contigs.csv",sep=",",header=TRUE)
dCDS <- read.table(file="Ddis/Reference_Based_DictyCDS-2Sep2014.fas_contigs.csv",sep=",",header=TRUE)
pasa1 <- read.table(file="Ddis/Reference_Based_pasa_dicty_Alt_Para_Annot.assemblies.fasta_contigs.csv",sep=",",header=TRUE)
pasa2 <- read.table(file="Ddis/Reference_Based_PASA_update_transcripts.fasta_contigs.csv",sep=",",header=TRUE)

d_frame <- data.frame(D_Discoideum="DictyCDS",Score=dCDS$score)
t_frame <- data.frame(D_Discoideum="TrinityAssembly",Score=trinity$score)
p1_frame <- data.frame(D_Discoideum="PASA1",Score=pasa1$score)
p2_frame <- data.frame(D_Discoideum="PASA2",Score=pasa2$score)

all_frame <- rbind(t_frame,p1_frame,p2_frame,d_frame)
library(ggplot2)
library(scales)
p <- ggplot(all_frame,aes(x=D_Discoideum,y=Score,fill=D_Discoideum))+geom_boxplot()+xlab("")+ggtitle("Dictyostelium discoideum")+ ylab("Score")+theme_bw() +theme(axis.line = element_line(colour = "black"),panel.grid.major = element_blank(),panel.grid.minor = element_blank(),panel.border = element_blank(),panel.background = element_blank(),text = element_text (size=15) , legend.position ='none')

Polysphondylium Pallidum

trinity <- read.table(file="Ppal/Reference_Based_NewAssembly_35631.fasta.clean_contigs.csv",sep=",",header=TRUE)
dCDS <- read.table(file="Ppal/Reference_Based_DNA.fas_contigs.csv",sep=",",header=TRUE)
pasa1 <- read.table(file="Ppal/Reference_Based_pasa_Pallidum_Gernot.assemblies.fasta_contigs.csv",sep=",",header=TRUE)
pasa2 <- read.table(file="Ppal/Reference_Based_PASA2.fasta_contigs.csv",sep=",",header=TRUE)

d_frame <- data.frame(P_Pallidum="PpalCDS",Score=dCDS$score)
t_frame <- data.frame(P_Pallidum="TrinityAssembly",Score=trinity$score)
p1_frame <- data.frame(P_Pallidum="PASA1", Score=pasa1$score)
p2_frame <- data.frame(P_Pallidum="PASA2", Score=pasa2$score)

all_frame <- rbind(t_frame,p1_frame,p2_frame,d_frame)
p1 <- ggplot(all_frame,aes(x=P_Pallidum,y=Score,fill=P_Pallidum))+geom_boxplot()+xlab("")+ggtitle("Polysphondylium pallidum")+ ylab("Score")+theme_bw() +theme(axis.line = element_line(colour = "black"),panel.grid.major = element_blank(),panel.grid.minor = element_blank(),panel.border = element_blank(),panel.background = element_blank(),text = element_text (size=15), legend.position ='none')

Dictyostelium Fasciculatum

trinity <- read.table(file="Dfas/Reference_Based_Trinity_blobplot_filter.fasta_contigs.csv",sep=",",header=TRUE)
dCDS <- read.table(file="Dfas/Reference_Based_DNA.fasta_contigs.csv",sep=",",header=TRUE)
pasa1 <- read.table(file="Dfas/Reference_Based_pasa_Dictyostelium_Fasciculatum_Assembly.assemblies.fasta_contigs.csv",sep=",",header=TRUE)
pasa2 <- read.table(file="Dfas/Reference_Based_PASA2.fasta_contigs.csv",sep=",",header=TRUE)

d_frame <- data.frame(D_Fasciculatum="DFasCDS",Score=dCDS$score)
t_frame <- data.frame(D_Fasciculatum="TrinityAssembly",Score=trinity$score)
p1_frame <- data.frame(D_Fasciculatum="PASA1",Score=pasa1$score)
p2_frame <- data.frame(D_Fasciculatum="PASA2",Score=pasa2$score)

all_frame <- rbind(t_frame,p1_frame,p2_frame,d_frame)
p2 <- ggplot(all_frame,aes(x=D_Fasciculatum,y=Score,fill=D_Fasciculatum))+geom_boxplot()+xlab("")+ggtitle("Dictyostelium fasciculatum")+ylab("Score")+theme_bw() +theme(axis.line = element_line(colour = "black"),panel.grid.major = element_blank() ,panel.grid.minor = element_blank(),panel.border = element_blank(),panel.background = element_blank(),text = element_text (size=15),legend.position='none')

Dictyostelium Lacteum

trinity <- read.table(file="Dlac/Reference_Based_Trinity_blobplot_filter.fasta_contigs.csv",sep=",",header=TRUE)
dCDS <- read.table(file="Dlac/Reference_Based_DNA.fas_contigs.csv",sep=",",header=TRUE)
pasa1 <- read.table(file="Dlac/Reference_Based_pasa_Dictyostelium_Lacteum_Assembly.assemblies.fasta_contigs.csv",sep=",",header=TRUE)
pasa2 <- read.table(file="Dlac/Reference_Based_PASA2.fasta_contigs.csv",sep=",",header=TRUE)

d_frame <- data.frame(D_Lacteum="DLacCDS",Score=dCDS$score)
t_frame <- data.frame(D_Lacteum="TrinityAssembly",Score=trinity$score)
p1_frame <- data.frame(D_Lacteum="PASA1",Score=pasa1$score)
p2_frame <- data.frame(D_Lacteum="PASA2",Score=pasa2$score)

all_frame <- rbind(t_frame,p1_frame,p2_frame,d_frame)
p3 <- ggplot(all_frame,aes(x=D_Lacteum,y=Score,fill=D_Lacteum))+geom_boxplot()+xlab("")+ggtitle("Dictyostelium lacteum")+ylab("Score")+theme_bw() +theme(axis.line = element_line(colour = "black"),panel.grid.major = element_blank(), panel.grid.minor = element_blank(),panel.border = element_blank(),panel.background = element_blank(),text = element_text (size=15),legend.position='none')

Combine All plots

source("../multiplot.R")
pdf("Contig_Score.pdf",width=15,height=15)
multiplot(p,p2,p1,p3)
dev.off()

Publication

Singh R, Lawal HM, Schilde C, Glöckner G, Barton GJ, Schaap P, Cole C. Improved annotation with de novo transcriptome assembly in four social amoeba species. BMC genomics. 2017 Dec 1;18(1):120.