Daniel:Notebook/ComboLock/2016-9-27: Difference between revisions

From ZhangLabWiki
Jump to navigation Jump to search
>Djacobse
>Djacobse
 
(One intermediate revision by the same user not shown)
Line 84: Line 84:
===Methods-Blast===
===Methods-Blast===


To do blast, I used a database made of the various primer, adapter, and index sequences. I set a threshold of 80% match to get rid of small alignments (that is, the read must contain >=80% of the bases matching to the primer/adapter/index). The table below summarizes the sequences targeted and how many bp a match requires.
To do blast, I used a database made of the various primer, adapter, and index sequences. I set a threshold of 80% alignment length match to get rid of small alignments (that is, the read must contain >=80% of the bases matching to the primer/adapter/index). The table below summarizes the sequences targeted and how many bp a match requires. [[Media:Analysis.miseq20160909.txt|Shell script for blast analysis]]


{| class="wikitable" <hiddentext>generated with [[:de:Wikipedia:Helferlein/VBA-Macro for EXCEL tableconversion]] V1.8</hiddentext>
{| class="wikitable" <hiddentext>generated with [[:de:Wikipedia:Helferlein/VBA-Macro for EXCEL tableconversion]] V1.8</hiddentext>
Line 162: Line 162:
| align="center" align="center" valign="bottom" | 7
| align="center" align="center" valign="bottom" | 7


|}  
|}
 
===Interior Blast===
 
I also performed a blast on the templates in combolock (padlocks, positive control templates, and full sequencing primers) against the individual sequences, using the templates as the db. In the table below are the matches that were not expected (off-target matches). For this one I did not do any thresholding since I want to see all possible off-target bindings, although blast seems to have a threshold of at least 7 bases. Highlighted in the table below are the most concerning sequence matches (it's all the same similarity). That is the LinkerR region matches to somewhere on Primer2. It is a 9bp match, and while it is not very likely, is certainly possible to cause problems.
 
blastn -db combolockseq.fa -query primerseqs.fa -task blastn-short -outfmt 7
 
{| class="wikitable" <hiddentext>generated with [[:de:Wikipedia:Helferlein/VBA-Macro for EXCEL tableconversion]] V1.8</hiddentext>
|- style="background-color:#B7DEE8;font-size:12pt;font-weight:bold" align="center"
| width="127" height="30" | query id
| width="98" |  subject id
| width="65" |  % identity
| width="64" |  alignment length
| width="70" |  mismatches
| width="65" |  gap opens
| width="65" |  q. start
| width="65" |  q. end
| width="65" |  s. start
| width="65" |  s. end
| width="65" |  evalue
| width="65" |  bit score
| width="89" | Notes
 
|- style="background-color:#FFFF00;font-size:12pt"  valign="bottom"
| height="15" | primer2
| Padlock0001-68
| align="center" | 100
| align="center" | 9
| align="center" | 0
| align="center" | 0
| align="center" | 5
| align="center" | 13
| align="center" | 84
| align="center" | 76
| align="center" | 0.014
| align="center" | 18.3
| LinkerR*
 
|- style="background-color:#FFFF00;font-size:12pt"  valign="bottom"
| height="15" | primer2
| AmpF6.4
| align="center" | 100
| align="center" | 9
| align="center" | 0
| align="center" | 0
| align="center" | 5
| align="center" | 13
| align="center" | 54
| align="center" | 46
| align="center" | 0.014
| align="center" | 18.3
| LinkerR*
 
|- style="font-size:12pt"  valign="bottom"
| height="15" | primer6
| Amp3
| align="center" | 100
| align="center" | 8
| align="center" | 0
| align="center" | 0
| align="center" | 4
| align="center" | 11
| align="center" | 44
| align="center" | 51
| align="center" | 0.057
| align="center" | 16.4
| L3BC*
 
|- style="background-color:#D9D9D9;font-size:12pt"  valign="bottom"
| height="15" | IlluminaP5
| Amp3
| align="center" | 100
| align="center" | 7
| align="center" | 0
| align="center" | 0
| align="center" | 4
| align="center" | 10
| align="center" | 61
| align="center" | 55
| align="center" | 0.26
| align="center" | 14.4
| Primer6*
 
|- style="font-size:12pt"  valign="bottom"
| height="15" | IlluminaP5
| Amp1
| align="center" | 100
| align="center" | 7
| align="center" | 0
| align="center" | 0
| align="center" | 4
| align="center" | 10
| align="center" | 61
| align="center" | 55
| align="center" | 0.26
| align="center" | 14.4
| Primer6*
 
|- style="background-color:#FFFF00;font-size:12pt"  valign="bottom"
| height="15" | LinkerR
| Amp3
| align="center" | 100
| align="center" | 9
| align="center" | 0
| align="center" | 0
| align="center" | 15
| align="center" | 23
| align="center" | 13
| align="center" | 5
| align="center" | 0.023
| align="center" | 18.3
| Primer2*
 
|- style="background-color:#FFFF00;font-size:12pt"  valign="bottom"
| height="15" | LinkerR
| Amp1
| align="center" | 100
| align="center" | 9
| align="center" | 0
| align="center" | 0
| align="center" | 15
| align="center" | 23
| align="center" | 13
| align="center" | 5
| align="center" | 0.023
| align="center" | 18.3
| Primer2*
 
|- style="font-size:12pt"  valign="bottom"
| height="15" | Latch3Barcode
| Amp3
| align="center" | 100
| align="center" | 7
| align="center" | 0
| align="center" | 0
| align="center" | 2
| align="center" | 8
| align="center" | 71
| align="center" | 65
| align="center" | 0.057
| align="center" | 14.4
| Primer6*
 
|- style="background-color:#D9D9D9;font-size:12pt"  valign="bottom"
| height="15" | Latch3Barcode
| Amp1
| align="center" | 100
| align="center" | 7
| align="center" | 0
| align="center" | 0
| align="center" | 2
| align="center" | 8
| align="center" | 71
| align="center" | 65
| align="center" | 0.057
| align="center" | 14.4
| Primer6*
 
|}
 
==Discussion==
 
So index34 (~130bp product) and index 36 (~100bp) product have similar content. For both, the R1 reads contain (~80% and ~70% respectively) the Illumina P7 and LinkerL regions, which are both part of the AmpR6.3 sequencing primer. The R2 reads contain (~50%) the Illumina P5 and LinkerR, which are the components of AmpF6.4Sol. This means that a good number of the reads are probably AmpF6.4 and AmpR6.3 dimers. I think the difference might be between primer dimers and off-target binding to primer 2.


[[Image:Miseq20160912-BlastCountsHistogram.png|800px]]
[[Image:Miseq20160912-BlastCountsHistogram.png|800px]]


[[Category:ComboLock]] [[Category:20160829]]
[[Category:ComboLock]] [[Category:20160829]]

Latest revision as of 21:30, 28 September 2016

Positive Control X3 (Started 08/29/2016)[edit]

Back to Calendar

Sequencing data in Genome miner at /media/NAS3_volume2/Dan/combolock/miseq_20160912

Data Analysis[edit]

As a reminder, here is a table of the three samples, as well as the size select gels for the library prep.

Sample Index Approx bp Size
Sample0 34 130
Sample1 35 >500
Sample2 36 100

Methods-Manual Inspection[edit]

I used several methods to analyze this data set. For the simplest, I just took the raw sequences and put them into excel charts. Example for one read below:

Name Read 1/2 Sequence Offset Seq 1 Seq1 Match Index34? (R1) Seq2 Seq 2 Match Seq3
@M00159:51:000000000-ATU5U:1:1101:15771:2370 1:N:0:AACTGA R1 GTGTTTGCGGCTAAGCGTTCCTAGCNANNGAATCTCGTATGCCGTCTTCTGCNNGCACAAAAAAACTCTTTTTNC GT GTTTGCGGCTAAGCGTTCCTAGC AmpR6.3-RevComp NANNGA ATCTCGTATGCCGTCTTCTGCNNGC AmpR6.3-RevComp bp: 33-53 ACAAAAAAACTCTTTTTNC
@M00159:51:000000000-ATU5U:1:1101:15771:2370 2:N:0:AACTGA R2 ACTTCGGACCTCGATAACATTTGAGAGTGGTGTAGATCTTGATGTTCGCCGAATCNTTAAAAANAAATGTNTTCNT AC TTCGGACCTCGATAACATTTGAG AmpF6.4-RevComp AGTGGT GTAGATCTTGATGTTCGCCGAATCN First 11 match AmpF6.4Rev Comp 30-40 TTAAAAANAAATGTNTTCN

These allow me to manually assess the sequences, but are also not great for larger scale processing (I only did ~4 for any given index). For larger scale analyses, I used blast.

Methods-Blast[edit]

To do blast, I used a database made of the various primer, adapter, and index sequences. I set a threshold of 80% alignment length match to get rid of small alignments (that is, the read must contain >=80% of the bases matching to the primer/adapter/index). The table below summarizes the sequences targeted and how many bp a match requires. Shell script for blast analysis

Name Length Number of bp Correct for 80% match
primer2 18 15
primer4 18 15
primer6 18 15
primer12 18 15
IlluminaP5 20 16
IlluminaP7 24 20
LinkerL 23 19
LinkerR 25 20
Latch1Barcode 8 7
Latch2Barcode 8 7
Latch3Barcode 8 7
Latch6Barcode 8 7
Latch8Barcode 8 7
Latch9Barcode 8 7

Interior Blast[edit]

I also performed a blast on the templates in combolock (padlocks, positive control templates, and full sequencing primers) against the individual sequences, using the templates as the db. In the table below are the matches that were not expected (off-target matches). For this one I did not do any thresholding since I want to see all possible off-target bindings, although blast seems to have a threshold of at least 7 bases. Highlighted in the table below are the most concerning sequence matches (it's all the same similarity). That is the LinkerR region matches to somewhere on Primer2. It is a 9bp match, and while it is not very likely, is certainly possible to cause problems.

blastn -db combolockseq.fa -query primerseqs.fa -task blastn-short -outfmt 7 
query id subject id % identity alignment length mismatches gap opens q. start q. end s. start s. end evalue bit score Notes
primer2 Padlock0001-68 100 9 0 0 5 13 84 76 0.014 18.3 LinkerR*
primer2 AmpF6.4 100 9 0 0 5 13 54 46 0.014 18.3 LinkerR*
primer6 Amp3 100 8 0 0 4 11 44 51 0.057 16.4 L3BC*
IlluminaP5 Amp3 100 7 0 0 4 10 61 55 0.26 14.4 Primer6*
IlluminaP5 Amp1 100 7 0 0 4 10 61 55 0.26 14.4 Primer6*
LinkerR Amp3 100 9 0 0 15 23 13 5 0.023 18.3 Primer2*
LinkerR Amp1 100 9 0 0 15 23 13 5 0.023 18.3 Primer2*
Latch3Barcode Amp3 100 7 0 0 2 8 71 65 0.057 14.4 Primer6*
Latch3Barcode Amp1 100 7 0 0 2 8 71 65 0.057 14.4 Primer6*

Discussion[edit]

So index34 (~130bp product) and index 36 (~100bp) product have similar content. For both, the R1 reads contain (~80% and ~70% respectively) the Illumina P7 and LinkerL regions, which are both part of the AmpR6.3 sequencing primer. The R2 reads contain (~50%) the Illumina P5 and LinkerR, which are the components of AmpF6.4Sol. This means that a good number of the reads are probably AmpF6.4 and AmpR6.3 dimers. I think the difference might be between primer dimers and off-target binding to primer 2.

File:Miseq20160912-BlastCountsHistogram.png