Chris:LabNotes/sci-Methyl Seq/Calendar/2017/2017-4-28

From ZhangLabWiki
Jump to navigation Jump to search

sci-Methyl Seq Barcode 2 Design[edit]

Background[edit]

5' ----------DDDDDDDD[Barcode2]-----------------NN 3'
   (Filler 3)  UMI             revComp(Filler 2)
  • There are several characteristics of Adapter 2 that we want to include:
    • Filler 3 = For second strand synthesis and PCR after bisulfite conversion in order to add P5/P7 adapter sequences. For this reason, we may want to use a 3-base nucleotide library for Filler 3 (A,G,T) in order to avoid potential errors in PCR due to incomplete bisulfite conversion
    • DDDDDDDD = 8-base UMI sequence using the 3-nucleotide library (D=A,G,T from standard IDT mixed base)
    • Barcode 2 = Likewise using only D's to avoid confusion after bisulfite conversion. This will serve as the second barcode added during the combinatorial barcoding steps in the protocol
    • revComp(Filler 2) = reverse complement of Filler 2 except for last two bases, which we want to mismatch
    • NN = This is a 2-base mismatch sequence used to prevent second strand synthesis in the other direction (we don't want to make a copy of the methylated C*'s in the sequence, so only want second strand synthesis to copy over Barcode 2, UMI and Filler 3 to the original template. NOTE: This will basically be the last two bases of Filler 2 but mismatched

Barcode 2 Adapter Design[edit]

  • We will derive Filler 3 using a similar method we used to derive Filler 2 using Dan's script (primergenerator.py). See <https://docs.google.com/document/d/1f6fUYy26eaqqJKdxxTBFgeVucB0uOuHPdfUKYiLDF9U/edit?usp=sharing>
    • Consequently, we want Filler 3 to have a Tm of ~58.8C-61.4C with a length of 20bp
    • However, unlike Filler 2, we want to make sure that Filler 3 does not contain any G's. Otherwise, the copied sequence when doing second strand synthesis would contain C's, which could potentially be converted during bisulfite conversion
    • Below is a table of potential Filler 3 sequences with the one we decided to use bolded (we want to also evaluate whether these sequences passed the self-annealing and hairpin formation tests from <http://www.bioinformatics.org/sms2/pcr_primer_stats.html>)
CCCCTCCTCCCCTTTTCACC    57.9    pass (high GC content)
CCTCTCCTTACCTCCTCTCC    55.9    pass
AACCATCCCCCCCCTTCCTT    55.9    pass (warning, contains runs of C's)
TACCTCCACCTCTCCCCCAC    57.9    pass (warning, contains runs of C's; high GC content; GC clamp warning)
CCTCTCCACCCCATCCTATC    55.9    pass
CCAAACCCCCCTTCTACCCT    55.9    pass (warning, contains runs of C's)
ACTACCCCTACTCTCCCCCC    57.9    pass (warning, contains runs of C's; high GC content; GC clamp warning)
ACTCTACCAACCCCCCCCAC    57.9    pass (warning, contains runs of C's; high GC content; GC clamp warning)
TACCCCCCCCCCCTCCAAAA    57.9    pass (warning, contains runs of C's; high GC content)
CCCCCCCCCACAAAATCTTC    55.9    pass (warning, contains runs of C's)
  • We will choose a random 8-base UMI and Barcode 2 to start with (make sure that these don't have any G's in them because they will be copied as C's in bisulfite conversion):
UMI = CCTCTCCA
Barcode 2 = CCTATC
  • The following is the final version of the test of Adapter 2:
5' CCTCTCCTTACCTCCTCTCCCCTCTCCA[CCTATC]AAACGCCGGGTAGGAGGGTG 3'
                                                         ^
                                                         |
                                                        Two base mismatch so not exactly reverse complementary to the original Filler 2 sequence