considerations - hints - solutions

2-Protein sequence homology search

• The protein sequence is shown in the EMBL entry annotation. You may use the EMBL entry as starting point to edit a protein sequence file that can then be reformatted into GCG format (sequence in gcg format and fasta format).
• Dotplot the query sequence against itself and against a high-scoring blast match in order to reveal its internal repeat structure.
• Split query sequence into repeat regions and unique parts: recommended partitioning: repeat region 1: pos. 1 - 313; repeat region 2: 314-502; unique region: 643-1433. You may use the GCG program ASSEMBLE to transfer these sequence regions to individual files.
• Using the the bioccelerator server one should know that the substitution matrices are scaled differently. For BLOSUM45 gap opening=15, and gap extension=2 may be reasonable choices.
• The only way to perform Smith-Waterman searches locally is by first constructing a profile from a single sequence and then doing a profilesearch against the protein database. The GCG recipe follows:

       profilemake -dat=genmoredata:blosum45.cmp -stringent <your querry sequence -outf=query.prf
       tprofilesearch -nosixframe -noave -normalize -minscore=5.0 -list=100 -gap=3.0 -len=0.3 -batch query.prf SW:*

The protein has the following internal repeat structure:

            200     400     600     800    1000    1200    1400 
              '       '       '       '       '       '       '
       [-R1-][-R1-][R2][R2][-R1-][--------unique region---------]

1. Only the fibronectin type 3 repeats. The N-terminal repeats may be homologous to a substrate-binding domain in SW:NANH_MICVI (Z-score=7).
2. At least three different types of domains: type 1 repeats, type 2 repeats (=FN3 domains), and a C-terminal unique region.
3. The protein is probably not a chitinase homolog because the similarity to chitinase is confined to the two FN3 domains, an extracellular module found in many different types of eukaryotic and prokaryotic proteins.

3 DNA sequence comparisons

• Use LOOKUP to find EMBL sequences. Compare promoter regions against REPBASE:* in order to find matches to common repetive elements.
• Use dotplots to identify multiple copies of the same type of repetitive elements.

See output files

The longest prometer regions are in EMBL entries HSENHREG1 (Z70243) and MMIL2R5 (M16398).

The mouse IL2Ra promoter region contains B1 and MT elements. The human IL2Ra promoter region contains Alu and Line-1 elements.

1. In the mouse promoter: -350 to transcription start, -1800 -1000.
2. Perhaps 25%.
3. The human Alu and mouse B1 elements are remotely related.
4. Available in EMBL are shorter upstream promoter regions from rat and from bovine.