MB620 Bioinformatics
University of New Haven
Instructor: Joel S. Bader
Class 6: Structure Databases. Sequence Assembly.

Agenda

• Protein structure databses: PDB, MMDB
• Sequence Assembly. Shotgun sequencing. Coverage. Reads, Contigs, Genes.

Structure Databases

• primary structure
• secondary structure (3 categories)
• tertiary structure (folds)
• quaternary structure

NMR vs. Xtal issues: resolution, biological conformation, sample preparation

How much detail do we retain?
residues
C-alpha trace
all atoms (most typical for structural databases)

How do we represent structural information?
labeled graph: atom types + positions + connections
minimal information: atom types + positions + rules for making connections

What does experimental data provide?
What about fluctuating structures?

PDB www.rcsb.org the Protein Data Bank, now at Rutgers

• Go to PDB
• Search for HIV protease
• Summary information
• View structure: stills, live action
• Display structure file: PDB, mmCIF format (macromolecular crystallographic info file)
• Structural neighbors (we'll get to this later)
• Other sources = db x-refs
• Sequence details = FASTA format, secondary structure summary

PDB file information

• HEADER with PDB-ID
• TITLE
• COMPND what's in the file, can have multiple cpds
• SOURCE, KEYWDS, EXPDTA (NMR, X-RAY), REVDAT, JRNL, REMARK are self-explanatory. Note the 6-letter names!
• DBREF database x-refs
• SEQRES explicit listing of the primary sequence information
• HELIX and SHEET information
• ATOM and HETATOM records for most of the atoms

MMDB http://www.ncbi.nlm.nih.gov/Structure/

• Search for HIV protease
• Notice the same records, link to PDB
• Sequence neighbors (pre-computed blast, more later)
• Structure neighbors (pre-computed)

Sequence neighbors vs. structure neighbors. Divergent vs. convergent evolution.

Viewing structures and structure neighbors

• Download and install Cn3D 2.0
• View a structure
• Go to structure neighbors
• How do you measure how close two structures are?
• Choose an alignment and view it

Sequence Assembly

Assembly algorithm

• Start with a sequence
• Build a cluster (sequence similarity to completion)
• Assemble
• Iterate

Example: building an assembly from ESTs

Take this sequence into CuraTools:

       1  cgagttcgtc aacgccgctt tcaacgtgac tgtggtggcc acaacacgtg tgggactccg
      61  cccgaggaat actgtgtgca gaccggggtg accgggtcac aagtcctgtc acctgtgcga
     121  cgccgggcag ccccacctgc agcacagggc agccttcctg accgactaca acaaccaggc
     181  cgacaccacc tggtggcaaa gcagagccat gct

Automated iteration: The Est Extractor at http://hercules.tigem.it/BLASTEXTRACT/estextract.html

Homework for Week 7

1. Hemoglobin and myoglobin are thought to have evolved from a common ancestor. Show an alignment of hemoglobin and myoglobin structures from the same species.
2. X-tal structures have been used to help develop better HIV protease inhibitors. Show an alignment of HIV protease with a peptide inhibitor and a non-peptide inhibitor.
3. Here is a fragment of DNA sequence:

           1 tcaggagcca gccccaccct tagaaaagat gttttccatg aggatcgtct gcctggtcct
          61 aagtgtggtg ggcacagcat ggactgcaga tagtggtgaa ggtgactttc tagctgangg
         121 aggaggcgtg cgtggcccaa gggttgtgga aagacatcaa tctgcctgca aagattcaga
         181 ctggcccttc tgctctgatg aagactggaa ctacaaatgc ccttctggct gcaggatgaa
         241 aagggttgat tgatgaagtc aatcaagatt ttacaaacag aataaataag ctcaaaaatt
         301 cactatttga atatcagaag ancaataagg attctcattc gttgaccact aatataatgg
         361 gaaattttga gaggcgattt ttcctcagcc aattaaccgt ggataatacc tacaaccgag
         421 tgtccagagg atctgaggan gcaggaattt gaagtcctga agcgcaaagt cataggaaaa
         481 gtncagcata tccagcttct ncagaaantg ttaggagctc ngtttggt
   

   See if you can assemble it with other ESTs to build a contig. How long is the contig?
4. Here is another DNA fragment:

           1 aaatttgtca tggatggagg gtatctggat caacccgaca actgtccaga gagagtcact
          61 gacctcatgc gcatgtgctg gcaattcaac cccaagatga ggccaacctt cctggagatt
         121 gtcaacctgc tcaaggacga cctgcacccc agctttccag aggtgtcgtt cttccacagc
         181 gaggagaaca aggctcccga gagtgaggag ctggagatgg agtttgagga catggagaat
         241 gtgcccctgg accgttcctc gcactgtcag agggaggagg cggggggccg ggatggaggg
         301 tcctcgctgg gtttcaagcg gctacgagga acacatccct tacacacaca tgaacggagg
         361 caagaaaaac gggcggattc tgaccttgcc tcggtccaat ccttcctaac agtgcctacc
         421 gtggcggggg cgggcagggg ttccattttc gctttcctct ggtttgaaag cctctggaaa
         481 actcaggatt ctcacgactc taccatgtcc aatggagttc agagatcgtt cctatacatt
         541 tctgttcatc ttaaggtgga ctcgtttggt taccaattta a
   

   See if you can build an EST assembly.
5. Take the contigs you built and now search against GenBank to find out what gene the ESTs are from. Which contig has more component sequences? Why do you think that is?
6. Blast the human genes against mouse ESTs. How does coverage of the two genes compare?
7. Bonus question.