Off topic:Trinity Reads are not idnetical
0
0
Entering edit mode
5.2 years ago

My reads names are not identical even after treating with "awk cmd "

here are the commands that I used

  awk '{{print (NR%4 == 1) ? "@1_" ++i "/1": $0}}'  Read_1.fastq > Raed_1_renamed.fastq
  awk '{{print (NR%4 == 1) ? "@2_" ++i "/2": $0}}'  Raed_2.fastq > Raed_2_renamed.fastq

after this I have perform the Triimomatic cmd for trimming but in the out put I got very less number of reverse read here is a cmd

java -jar /share/apps/Trimmomatic-0.38/trimmomatic-0.38.jar PE -phred33 Read_1_renamed.fastq Read_2_renamed.fastq  FP_Read_1.fastq FU_Read_1.fastq RP_Read_2.fastq  RU_Raed_2.fastq ILLUMINACLIP:/share/apps/Trimmomatic-0.38/adapters/TruSeq3-PE-2.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 HEADCROP:10 MINLEN:36

Output:-

Using PrefixPair: 'TACACTCTTTCCCTACACGACGCTCTTCCGATCT' and 'GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT'
Using Long Clipping Sequence: 'AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTA'
Using Long Clipping Sequence: 'AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC'
Using Long Clipping Sequence: 'GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT'
Using Long Clipping Sequence: 'TACACTCTTTCCCTACACGACGCTCTTCCGATCT'
ILLUMINACLIP: Using 1 prefix pairs, 4 forward/reverse sequences, 0 forward only sequences, 0 reverse only sequences
Input Read Pairs: ****47190247 Both Surviving: 44940671 (95.23%) Forward Only Surviving: 1504722 (3.19%) Reverse Only Surviving: 365037 (0.77%) Dropped: 379817 (0.80%)****
TrimmomaticPE: Completed successfully

after this the size of the file is

Read_1_renamed.fastq 9.7 GB Read_2_renamed.fastq 9.7 GB

FP_Read_1.fastq 8.3G FU_Read_1.fastq 253M RP_Read_2.fastq 8.3G RU_Raed_2.fastq 62M

After this I tried to run the trinity and it was aborted stating the following error

/share/apps/trinityrnaseq-Trinity-v2.8.3/Trinity --seqType fq --left FP_Read_1.fastq,FU_Read_1.fastq --right RP_Read_2.fastq,RU_Raed_2.fastq --CPU 10 --max_memory 96G

Error, pairs.K25.stats is empty.  Be sure to check your fastq reads and ensure that the read names are identical except for the /1 or /2 designation. at /share/apps/trinityrnaseq-Trinity-v2.8.3/util/insilico_read_normalization.pl line 921.

later i have checked tried to see the head so I found this

[sneharl@compute-0-8 nep11]$ head RP_Read_2.fastq 
@2_13/2
ACATAATCTCACTCGACGTACCAGGCATGAGGAGGGAGGATGTCAAGATAGAGGTGGAGGAGAACAGGGTGCTGAGGGTGAGCGGAGAGAG
+
FHHJJIIJJGIJJIJ?HIFFHJCHIIGDHBEH<DGA@FHCGC=AAEHEFDFFFE6=AABBB@?ADA?AB5?@CCCDDD088AA<<<958<<
@2_15/2
TCAATTCATCAACAACCTTAGATCTCAACTCATCAGCAAGCTCTTTCTGGGCAGTATCACTTGGAGATTTCTTCCCTGTCCTAAGACAAGC
+
HHHJJJJJIJJIJJIJJJJJJIIIIIJJJIGJIJJJJIJHHIIJIJJHIJJJFIDHIJJJJIJIJJHHHHHHFFFFFF;AECCEDDDDDDD
@2_16/2
GCAGAAACAGAAGGGTAATTCAGTGCTACTGCATCAAAGACTGTCTGCCTGACACAGTTGGGAGTTTGAACACCAACAATAACCAAGCTAT
[sneharl@compute-0-8 nep11]$ head RU_Raed_2.fastq 
@2_78/2
AAGCATCACGTCAAATGAACAGCCGTACAATACGCAGCGCACCTTATTCCAACGCCTTTTCTCGTCAACGATTTACGATTGCAAATTATCA
+
HHHJJJIJJJJJJJJJJIJJJJJJJIJIJIIJEHJJIJJIHIHHHHHHFFFFFDDDDDDDDDDDDDDDDDDDDDDEBDDD?CCDDDDDEDC
@2_682/2
AATACAAGAAAATTTCGTCTCATTCAAAAGTCCCTT
+
A?D<<<:AFF3<FEFI@+A8?4?E@ECFC3*1?**0
@2_735/2
ACTATGACAGATATCGATACCGATATTTTCATCCATCCACCGGACCCAAAATATACTACCAAAAAGGAAATGATTTCTCTTCACTTCGTTC

I have tried to check the file left.fa.K25.stats and right.fa.K25.stats

[sneharl@compute-0-8 tmp_normalized_reads]$ head  left.fa.K25.stats
273661  257154  60829.4 23.6548 1_5/1   thread:0
3393    3310.83 452.798 13.6763 1_6/1   thread:4
551157  378893  292184  77.1153 1_4/1   thread:2
247465  248536  12803.3 5.15147 1_1/1   thread:1
2550    2421.5  783.635 32.3616 1_3/1   thread:3
715265  769537  114199  14.8399 1_9/1   thread:2
50212   49096.3 8357.57 17.0228 1_8/1   thread:4
888228  835519  213339  25.5337 1_10/1  thread:1
913960  907977  68618   7.55724 1_11/1  thread:3
913960  908888  69092.1 7.60182 1_7/1   thread:0
[sneharl@compute-0-8 tmp_normalized_reads]$ head right.fa.K25.stats
281168  278462  7324.79 2.63045 2_1/2   thread:3
2232    2306    146.529 6.35423 2_3/2   thread:0
3746    3771.5  88.5557 2.34802 2_6/2   thread:2
332464  329362  16988.3 5.15796 2_5/2   thread:4
325543  322963  16191   5.01327 2_4/2   thread:1
920338  922948  10396.2 1.12641 2_7/2   thread:3
62646   62614.5 994.886 1.58891 2_8/2   thread:0
1138774 1.14581e+06     38648.5 3.37302 2_9/2   thread:2
800599  811590  45665.9 5.62673 2_10/2  thread:4
821097  820930  31654.2 3.8559  2_11/2  thread:1
[sneharl@compute-0-8 tmp_normalized_reads]$

Please help me in this regard

next-gen sequencing sequence • 1.0k views
ADD COMMENT
This thread is not open. No new answers may be added
Traffic: 3182 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6