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

From ZhangLabWiki
Jump to navigation Jump to search
>Djacobse
(Created page with "=Positive Control X3 (Started 08/29/2016)= Back to Calendar Sequencing data in Genome miner at /media/N...")
 
>Djacobse
Line 36: Line 36:
File:2016-09-05-X3Dilution-SizeSelect-After.png|After Image
File:2016-09-05-X3Dilution-SizeSelect-After.png|After Image
</gallery>
</gallery>
===Methods-Manual Inspection===
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:
{| class="wikitable" <hiddentext>generated with [[:de:Wikipedia:Helferlein/VBA-Macro for EXCEL tableconversion]] V1.8</hiddentext>
|- style="font-size:12pt"  valign="bottom"
| width="183" height="45" | @M00159:51:000000000-ATU5U:1:1101:15771:2370 1:N:0:AACTGA
| width="55" | R1
| width="276" | GTGTTTGCGGCTAAGCGTTCCTAGCNANNGAATCTCGTATGCCGTCTTCTGCNNGCACAAAAAAACTCTTTTTNC
| width="110" | GT
| width="103" | GTTTGCGGCTAAGCGTTCCTAGC
| width="90" | AmpR6.3-RevComp
| width="65" | NANNGA
| width="196" | ATCTCGTATGCCGTCTTCTGCNNGC
| width="120" | AmpR6.3-RevComp bp: 33-53
| width="120" | ACAAAAAAACTCTTT
| width="65" | TTNC
|- style="font-size:12pt"  valign="bottom"
| height="45" | @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
| TTAAAAANAAATGTN
| TTCNT
|}
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===
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.
{| 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="86" height="45" | Name
| width="65" | Length
| width="101" | Number of bp Correct for 80% match
|- style="font-size:12pt"
| height="15"  valign="bottom" | primer2
| align="center" align="center" valign="bottom" | 18
| align="center" align="center" valign="bottom" | 15
|- style="background-color:#D9D9D9;font-size:12pt"
| height="15"  valign="bottom" | primer4
| align="center" align="center" valign="bottom" | 18
| align="center" align="center" valign="bottom" | 15
|- style="font-size:12pt"
| height="15"  valign="bottom" | primer6
| align="center" align="center" valign="bottom" | 18
| align="center" align="center" valign="bottom" | 15
|- style="background-color:#D9D9D9;font-size:12pt"
| height="15"  valign="bottom" | primer12
| align="center" align="center" valign="bottom" | 18
| align="center" align="center" valign="bottom" | 15
|- style="font-size:12pt"
| height="15"  valign="bottom" | IlluminaP5
| align="center" align="center" valign="bottom" | 20
| align="center" align="center" valign="bottom" | 16
|- style="background-color:#D9D9D9;font-size:12pt"
| height="15"  valign="bottom" | IlluminaP7
| align="center" align="center" valign="bottom" | 24
| align="center" align="center" valign="bottom" | 20
|- style="font-size:12pt"
| height="15"  valign="bottom" | LinkerL
| align="center" align="center" valign="bottom" | 23
| align="center" align="center" valign="bottom" | 19
|- style="background-color:#D9D9D9;font-size:12pt"
| height="15"  valign="bottom" | LinkerR
| align="center" align="center" valign="bottom" | 25
| align="center" align="center" valign="bottom" | 20
|- style="font-size:12pt"
| height="15"  valign="bottom" | Latch1Barcode
| align="center" align="center" valign="bottom" | 8
| align="center" align="center" valign="bottom" | 7
|- style="background-color:#D9D9D9;font-size:12pt"
| height="15"  valign="bottom" | Latch2Barcode
| align="center" align="center" valign="bottom" | 8
| align="center" align="center" valign="bottom" | 7
|- style="font-size:12pt"
| height="15"  valign="bottom" | Latch3Barcode
| align="center" align="center" valign="bottom" | 8
| align="center" align="center" valign="bottom" | 7
|- style="background-color:#D9D9D9;font-size:12pt"
| height="15"  valign="bottom" | Latch6Barcode
| align="center" align="center" valign="bottom" | 8
| align="center" align="center" valign="bottom" | 7
|- style="font-size:12pt"
| height="15"  valign="bottom" | Latch8Barcode
| align="center" align="center" valign="bottom" | 8
| align="center" align="center" valign="bottom" | 7
|- style="background-color:#D9D9D9;font-size:12pt"
| height="15"  valign="bottom" | Latch9Barcode
| align="center" align="center" valign="bottom" | 8
| align="center" align="center" valign="bottom" | 7
|}
[[Category:ComboLock]] [[Category:20160829]]

Revision as of 19:16, 28 September 2016

Positive Control X3 (Started 08/29/2016)

Back to Calendar

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

Data Analysis

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

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:

@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 ACAAAAAAACTCTTT TTNC
@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 TTAAAAANAAATGTN TTCNT

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

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.

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