1. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Joining and Rotating Data with Molecules Masanori Arita, Masami Hagiya and Akira Suyama Summarized by Sung-Kyu Kim

2. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Introduction zMolecular operation is so unreliable to perform sophisticated computations zFeasible way of implementing significant operations such as Cartesian product or selection in database theory zData encoded in the following form (tag data tag)+ zDNA manipulation using PCR

3. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Computer Database zTuples zOperations Selection, Projection, Union, Difference, Catesian product (or join) StudentIDName S0001Tom S0002July StudentIDSubjectScore S0001Math100 S0002English70 S0002Math60 tuples

4. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Performed Experiments zSimple polymerization zPCR from single-stranded templates zCircularization in diluted solution zUsing biotin for circularization Streptavidin-coated magnetic beads zExperiment output is analyzed by capillary electrophoresis system with LIFluor daDNA1000Kit and System Gold Output graph with RFU (relative fluorescence unit) measures the amount of dsDNA

5. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Sequence Design conditions zBoth 5bp ends of each data (or tag) do not appear in the ends of other data zThe GC content of each data (or tag) is not greatly biased zA tag sequence does not form a stable structure by itself zssDNA for generating each tuple anneals at the expected sites

6. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ zEach tuple was made from ssDNA of 60 or 45 bp zPrimer sequence can be either 15 or 30 bp zDNA sequences for the experiments Data Representation (1/3)

7. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Data Representation (2/3) zData: striped block zTag: shaded block indicates what kind of data is stored after or before it Can be used as target sites for PCR primers

8. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Data Representation (3/3) zConcatenated sequence of seq1, seq2 and seq3 1-2-3 zssDNA has a header (fwd or rev) before a sequence of numbers Fwd: 5’ → 3’ Rev: 3’ → 5’ e.g) fwd 1-2-3-4, rev 4-5-6 zdsDNA has no header

9. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Simple polymerization zUsing overlapping region which anneals at low temperature zNot a good result RFU value was less than 20 Time consuming

10. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ PCR from single-stranded templates zDirect amplification of dsDNA from the ssDNA templates

11. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Concatenation of overlapping sequences zConcatenation of three templates tupleA(B) + seq7 seq6 + tupleC(D) + seq13 seq12 + tupleE With primers fwd1, rev18 Output: A-C-E or A-D-E or B-C-E or …

12. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ zUsing bridge primers Concatenation with bridges

13. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Rotation

14. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ zDNA favors circularization rather than polymerization in diluted solution zinverse PCR on circular DNA produces rotated data Rotation in diluted solution unexpected ligation

15. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ Rotation with biotin zUse biotin to avoid unexpected ligation

16. © 2004, SNU Biointelligence Lab, http://bi.snu.ac.kr/ zTuple management using PCR technique zReliable realization of catesian product operation zThis method can be useful for building initial population of PLM library and updating library Discussion