Trinity Assembly

[Bang_Didi]

Dear All..
As a newbie in transcriptome analysis, I would like to ask a question about doing whole transcriptome assembly using trinity.

Is it possible for us to get two different transcripts when we assemble our reads by either concatenating the reads or listed the reads (using comma separation as Trinity manual says)?. I am just not sure with the results that I got, It seems that I got different transcripts (can tell this from its size which is different) using these two different method in preparing my reads for the assembly using Trinity. Any thought what went wrong ?

Cheers
Didi

[westerman]

Trinity is non-deterministic thus some variation between runs of it are expected. Not a lot but some.

[Bang_Didi]

Thanks for that westerman... Should I worry that the variation will also significantly be expressed when I construct the metrics for the transcripts evaluation?

[Bang_Didi]

FYI:

The Trinity stats that I got for the transcript that was built from concatenated data:
################################
## Counts of transcripts, etc.
################################
Total trinity 'genes': 236322
Total trinity transcripts: 518647
Percent GC: 45.98

########################################
Stats based on ALL transcript contigs:
########################################

Contig N10: 8296
Contig N20: 6856
Contig N30: 5744
Contig N40: 4826
Contig N50: 4031

Median contig length: 1217
Average contig: 2100.35
Total assembled bases: 1,089,337,664


#####################################################
## Stats based on ONLY LONGEST ISOFORM per 'GENE':
#####################################################

Contig N10: 6119
Contig N20: 4351
Contig N30: 3248
Contig N40: 2367
Contig N50: 1635

Median contig length: 367
Average contig: 799.05
Total assembled bases: 188,834,004

The Trinity stats that I got for the transcript that was built from listing all of the reads using comma separation:

################################
## Counts of transcripts, etc.
################################
Total trinity 'genes': 244,160
Total trinity transcripts: 301,140
Percent GC: 44.75

########################################
Stats based on ALL transcript contigs:
########################################

Contig N10: 6864
Contig N20: 5185
Contig N30: 4130
Contig N40: 3303
Contig N50: 2581

Median contig length: 448
Average contig: 1115.03
Total assembled bases: 335,781,132


#####################################################
## Stats based on ONLY LONGEST ISOFORM per 'GENE':
#####################################################

Contig N10: 5852
Contig N20: 4184
Contig N30: 3122
Contig N40: 2305
Contig N50: 1623

Median contig length: 374
Average contig: 806.19
Total assembled bases: 196,840,230

[westerman]

Those variations are more than I would expect and I can see why you are concerned. I'll see if I can fire up a recent Trinity assembly (I almost always use comma separated files) with combined reads and see what differences I get.

[ltutar]

Dear Bang_Didi,

Did you make a decision which way is the best comma separation or combining?

FYI:

The Trinity stats that I got for the transcript that was built from concatenated data:
################################
## Counts of transcripts, etc.
################################
Total trinity 'genes': 236322
Total trinity transcripts: 518647
Percent GC: 45.98

########################################
Stats based on ALL transcript contigs:
########################################

Contig N10: 8296
Contig N20: 6856
Contig N30: 5744
Contig N40: 4826
Contig N50: 4031

Median contig length: 1217
Average contig: 2100.35
Total assembled bases: 1,089,337,664


#####################################################
## Stats based on ONLY LONGEST ISOFORM per 'GENE':
#####################################################

Contig N10: 6119
Contig N20: 4351
Contig N30: 3248
Contig N40: 2367
Contig N50: 1635

Median contig length: 367
Average contig: 799.05
Total assembled bases: 188,834,004

The Trinity stats that I got for the transcript that was built from listing all of the reads using comma separation:

################################
## Counts of transcripts, etc.
################################
Total trinity 'genes': 244,160
Total trinity transcripts: 301,140
Percent GC: 44.75

########################################
Stats based on ALL transcript contigs:
########################################

Contig N10: 6864
Contig N20: 5185
Contig N30: 4130
Contig N40: 3303
Contig N50: 2581

Median contig length: 448
Average contig: 1115.03
Total assembled bases: 335,781,132


#####################################################
## Stats based on ONLY LONGEST ISOFORM per 'GENE':
#####################################################

Contig N10: 5852
Contig N20: 4184
Contig N30: 3122
Contig N40: 2305
Contig N50: 1623

Median contig length: 374
Average contig: 806.19
Total assembled bases: 196,840,230

[Nanu]

Greetings to all!

I would like to know about the reads/kmers per transcripts. As the TrinityStats.pl tells the total assembled bases. contig length. no . of transcripts as longest isoform. So I would like to know about the difference between Trinity.fasta and single.fasta.
When we execute the TrinityStats.pl , we know about the
1. Stats based on ONLY LONGEST ISOFORM per 'GENE
2.Stats based on ALL transcript contigs

May i know that Trinity.fasta contains all transcripts or it has genes also. ?