Version : 971103
Author : Gerard J. Kleywegt, Dept. of Cell and Molecular Biology,
Uppsala University, Biomedical Centre, Box 590,
SE-751 24 Uppsala, SWEDEN
E-mail : gerard@xray.bmc.uu.se
Purpose : introduction to the SBIN package
Package : SBIN
Reference(s) for the SBIN programs:
* 1 * G.J. Kleywegt & T.A. Jones (1998). Databases in protein crystallography. Acta Cryst D54, 1119-1131. [http://alpha2.bmc.uu.se/gerard/papers/databases.html] [http://www.iucr.org/iucr-top/journals/acta/tocs/actad/1998/actad5406_1.html]
* 2 * G.J. Kleywegt & T.A. Jones (1999 ?). Chapter 25.2.6. O and associated programs. Int. Tables for Crystallography, Volume F. To be published.
971024 - 0.1 - first version
971103 - 0.2 - expanded
SBIN ("Structural Bio-INformatics") is a package of programs that operate at the interface of structural biology and the world of protein sequences. Its major purpose is to help structural biologists to use the results of their structural studies in order to detect other proteins which may be related in structure and/or function.
In order to scan sequence profiles against SWISS-PROT, you will also need:
(1) the "pftools" suite of programs, written by Philipp Bucher ( mailto:pbucher@isrec-sun1.unil.ch ) and available by ftp from http://ulrec3.unil.ch:80/ftp-server/pftools/ (the suite should compile on most Unix machines).
(2) the SWISS-PROT database of protein sequences ( http://www.expasy.ch/sprot/sprot-top.html ), which can be downloaded by ftp from ftp://ftp.expasy.ch/databases/swiss-prot/ (at the time of writing, the file "compressed/sprot35.dat.Z").
If you only want to scan PROSITE ( http://www.expasy.ch/sprot/prosite.html ) patterns against the SWISS-PROT (and TREMBL) database, you don't need a local copy of the SWISS-PROT database or the pftools programs. Instead, you can use the WWW-based PROSITE server at http://www.expasy.ch/sprot/scnpsit2.html at ExPASy in Geneva.
Suppose that you solve a new protein structure which turns out to contain a fold which is (partly) similar to that of one or more other proteins (e.g., using DEJAVU or SPASM). If you align the two structures (e.g., using LSQMAN), you can feed them into the program STRUPAT which will look for more or less conserved residues in structurally conserved regions. It will use these to generate PROSITE-type sequence patterns (a.k.a. footprints, fingerprints, motifs, ...).
Such a pattern may look as follows: G-x(3)-C-x(2)-[ILV]. This means: glycine - three residues of any type - cysteine - two residues of any type - one residue of type Ile/Leu/Val. A protein which contains the peptide GYAVCPSV would fit this pattern.
STRUPAT requires as input (sensible defaults are provided):
- a random number seed (to generate a random sequence to see how
informative any generated patterns are, i.e. how often they
occur by chance)
- a distance cut-off to define core equivalent residues in the
aligned structures
- a distance cut-off to extent the structurally equivalent core
with residues which are a bit further apart
- whether or not the program should try to correct "frame-shifts"
in the aligned structures
- minimum number of residues in any generated patterns
- parameters determining when 2, 3 or 4 different residue types
should be used in the pattern
- name of the PDB file which contains the *superimposed* structures
Example output (when STRUPAT encounters a sufficiently long, structurally conserved stretch) for a few aligned lipocalin models:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- New structurally conserved stretch ! Starts at residue ASN - 14 molecule 2 @ ASN - 14 molecule 3 @ ASP - 5 molecule 4 @ ASN - 17 molecule 5 @ ASN - 13 |NNDNN| ==> [ND] - |FFFFF| ==> F - |DDDDN| ==> [DN] - |KKIWV| ==> X - |AASSE| ==> X - |RRKNK| ==> X - |FFFYI| ==> X - |SALHN| ==> X - |GGGGG| ==> G - |TTFKE| ==> X - |WWWWW| ==> W - |YYYWH| ==> X - |AAEET| ==> X - |MMIVI| ==> X - |AAAAI| ==> [AI] - |KKFKL| ==> X - |KKAYA| ==> X - Nr of residues : ( 17) RMS (RMSD) (A) : ( 0.828) PROSITE pattern : ([ND] - F - [DN] - X(5) - G - X - W - X(3) - [AI]) Length : ( 15) "Information" : ( 7.577) Nr of matches to random sequence : ( 0) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Running this pattern on the PROSITE server gives a set of 29 matches, all of which are lipocalins. However, this is only a subset of all lipocalin sequences in the database. The reason why the others are not picked up is that there were too few sequence in the input which leads to too strict a pattern (e.g., in some sequences the required tryptophan is actually a tyrosine).
The program also found another pattern, namely: [IV]-X(2)-T-D-[YN]-X-[TN]-[FY]. This pattern generates 39 matches, but this includes some false hits. False hits occur when the pattern is too weak/general/aspecific.
If we combine the two patterns, allowing 50-150 residues in between them (indicated by "-x(50,150)-" in the pattern), 20 matches are found, all of which are lipocalins (of course, since they all match the first pattern which only yielded lipocalins as hits).
The following four numbers are relevant when judging the quality
of a pattern:
- true positives: matches which are really hits
- true negatives: non-matching sequences which are not hits
- false positives: matching sequences which are not hits
- false negatives: real hits which do not match the pattern
Two simple statistics can be used to capture the database retrieval
quality:
- recall = #truepos / (#truepos + #falseneg)
- reliability = #truepos / (#truepos + #falsepos)
Recall is the fraction of real hits which match the pattern (between 0 and 1, the higher the better); reliability is the fraction of patterns which match the sequence and which are real hits (also between 0 and 1, the higher the better).
For an ideal pattern, the number of false positives and negatives is zero. Only real hits match the pattern, and all sequences that match the pattern are real hits. In that case, both recall and reliability are one. Longer patterns, with more highly conserved residues are likely to have better reliability, whereas shorter and/or less specific patterns will have better recall.
A more sophisticated technique to find related sequences is the use of profiles. A profile is a matrix where every residue has a row of numbers associated with it, which indicate how well each of the twenty residue types "fit in" at that particular position in the sequence. For instance:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Gly Ala Ser ... Phe Tyr Trp ...
...
Ala 263 2 5 3 -2 -2 -4
Phe 264 -4 -3 -4 10 9 7
...
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
A profile can be aligned with all sequences in a database. A sequence which is "compatible" with the profile will receive a high score. For instance, a sequence containing the dipeptide Ala-Tyr would obtain a score of 5 + 9 = 14; Tyr-Ala, on the other hand, would only score -2 + -3 = -5.
Note that, in addition to the twenty values for each of the common amino-acid residue types, the matrix also contains two columns which contain a score (or penalty) for the opening and extension of a gap (in the alignment of a database sequence to the profile).
Whereas patterns are very strict (if one strictly conserved residue is not conserved in one sequence, this sequence will not be matched to the pattern, even if it satisfies the rest of the pattern), profiles are more tolerant/subtle. For instance, if a residue is a tyrosine in all known sequences, a related sequence which happens to have a phenylalanine in that position may still obtain a high score.
Traditionally, profiles have been generated from multiple aligned
sequences. The actual values in the profile matrix depend on
three factors:
- the variety of residues observed in each position in the aligned
sequences (e.g., a strictly conserved Trp will lead to a high
value for the Trp-entry in that row of the matrix)
- knowledge about the likelihood of residue substitutions (e.g.,
Phe and Tyr are closely related residues, so a strictly conserved
Phe will also give a fairly high value for a Tyr in that position).
This knowledge is encoded in residue substitution tables (e.g.,
PAM and BLOSUM matrices)
- weights assigned to the individual sequences in the alignment to
reduce the effect of sample bias. For instance, if three sequences
AAAA, AAAA, and GGGG are used to generate a profile, the first
two are redundant and should receive a weight of 1/4 each, whereas
the third should be weighted by 1/2.
The program STRUPRO takes a slightly different approach. It takes as input a set of superimposed *structures*, and generates a profile only for stretches of residues that are in structurally equivalent positions. Inside such stretches, insertions are strongly penalised; in between insertions are "cost-neutral". The rationale is that, since structure is generally better conserved than sequence, a profile based only on the structurally-conserved core of a set of proteins stands a better chance of picking up other proteins from the database with a similar structure.
The profile can then be scanned against SWISS-PROT to reveal more proteins that could belong to the same class (structurally, functionally, evolutionarily).
A profile can be refined in an iterative process. For example, the program PRF2MSEQ can be used to produce a multiple-sequence alignment of the matches found by scanning SWISS-PROT with a profile. This alignment can subsequently be fed into MSEQPRO, which does the same as STRUPRO, but using aligned sequences (rather than 3D structures) as input. Like STRUPRO, MSEQPRO uses a conservative approach in that it only generates bits of profile for stretches of residues without any insertions and deletions in the multiple alignment.
Profile analysis can proceed in myriad ways. At the outset one needs either a set of aligned sequences, or a set of related and superimposed structures. If one solves a structure which does not look like any other structure in the PDB, one could start instead with only those residues which occur in secondary structure elements; generate a profile; scan SWISS-PROT; add the sequences of encouraging matches; generate a new profile, etc.
Below is a simple flow-chart. A hash "#" indicates a program from the SBIN package; a dollar "$" a program of the pftools package, and an ampersand "&" indicates a Unix command. The bottom part of the chart (below "happy ?") deals with running the profile against the entire SWISS-PROT database in order to obtain statistics to determine normalised scores as well as cut-offs. These may be relevant if you want to generate a library of profiles, or if you want to add them to PROSITE's set of profiles.
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----one or more structures one or more sequences | | V V STRUPRO # <----- substitution matrix ----> MSEQPRO # <---- | | | -----------> profile <---------------------- | | | V | pfsearch -ry $ | | | V | matching sequences | | | V | happy ? --> NO --> PRF2MSEQ # --> aligned sequences | V YES | V pfsearch -a $ | V sort -nr & | V scores for all proteins in SWISS-PROT | | V V pfscale $ ZPROF # | | V V score normalisation Z-score normalisation | | ------------ | V final profile
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Example commands:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- STRUPRO (interactive) pfsearch -ry pfx.prf /nfs/scr_uu5/gerard/sprot34.dat | & tee pfx.hits PRF2MSEQ < pfx.hits > pfx.seq MSEQPRO (interactive) pfsearch -a aligned.prf /nfs/scr_uu5/gerard/sprot34.dat | & tee pfsearch_all.log sort -nr pfsearch_all.log > pfsearch_all.sorted ; rm pfsearch_all.log pfscale pfsearch_all.sorted > pfsearch.scale ZPROF < pfsearch_all.sorted |& tee zprof.top ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
A profile file may look as follows:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ID MSEQPRO; MATRIX. AC PS99999; DT JAN-1900 (CREATED); DE Created by MSEQPRO V. 971023/0.3 at Thu Oct 23 21:51:45 1997 for user gerard CC CC Substitution matrix file : ../strupro_blosum45.lib CC Nr of sequences used : 33 CC Min fragment length : 5 CC Weighting scheme : S CC MA /GENERAL_SPEC: ALPHABET='ARNDCEQGHILKMFPSTWYV'; LENGTH= 56; MA TOPOLOGY=LINEAR; MA /DISJOINT: DEFINITION=PROTECT; N1=1; N2= 56; MA /CUT_OFF: LEVEL=0; SCORE= 500; MA /DEFAULT: MI=-100; I=-10; IM=0 ; MD=-100; D=-3; DM=0; MA /M: SY='E'; M=3,2,7,3,-20,12,4,-9,-2,-21,-23,3,-13,-26,-12,12,4,-29,-15,-17; MA /M: SY='R'; M=-9,31,1,-4,-27,9,7,-18,-6,-24,-20,28,-9,-24,-14,-8,-9,-22,-11,-18; MA /M: SY='Y'; M=-16,-18,-21,-30,-26,-21,-25,-32,-6,14,10,-21,6,33,-28,-20,-10,8,41,6; MA /M: SY='A'; M=9,-12,3,-11,-16,-7,-9,-10,-5,-6,-2,-13,-2,-11,-18,3,1,-28,-11,-6; MA /M: SY='G'; M=0,-20,0,-10,-30,-20,-20,70,-20,-40,-30,-20,-20,-30,-20,0,-20,-20,-30,-30; MA /M: SY='R'; M=-9,22,-3,-5,-26,7,1,-20,-6,-20,-16,17,-8,-17,-16,-6,-4,-18,-4,-15; MA /M: SY='W'; M=-20,-20,-40,-40,-50,-20,-30,-20,-30,-20,-20,-20,-20,10,-30,-40,-30,150,30,-30; MA /M: SY='Y'; M=-20,-11,-17,-22,-29,-13,-19,-28,22,-6,-2,-14,-1,30,-29,-20,-13,25,60,-12; MA /M: SY='E'; M=8,-6,-1,-6,-20,17,6,-9,-7,-15,-18,-3,-8,-26,-10,11,3,-26,-15,-14; MA /M: SY='I'; M=-6,-23,-20,-33,-23,-18,-26,-33,-23,33,14,-23,17,1,-22,-14,-4,-19,2,25; MA /M: SY='A'; M=24,-16,-10,-18,-17,-4,-10,-5,-10,-6,-6,-12,-4,-19,-15,1,-5,-21,-14,-2; MA /M: SY='K'; M=-10,7,-12,-14,-26,-2,-4,-25,-14,-6,-1,14,2,-12,-18,-16,-9,-19,-5,-4; MA /M: SY='A'; M=15,-10,-12,-20,-20,-7,-11,-16,-16,1,-3,-2,2,-12,-15,-6,-6,-17,-6,2; MA /M: SY='S'; M=-3,-12,4,12,-24,1,4,-4,-9,-26,-29,-7,-21,-29,13,14,2,-35,-22,-22; MA /M: SY='N'; M=-8,-9,13,8,-25,-6,-2,-12,-5,-16,-20,-5,-15,-19,5,0,-2,-30,-14,-19; MA /M: SY='E'; M=0,2,-2,-7,-21,17,3,-15,-6,-14,-17,7,-1,-24,-12,6,2,-26,-12,-11; MA /I: MI=0; I=-1; MD=0; /M: SY='X'; M=0; D=-1; MA /M: SY='Y'; M=-12,2,-9,-7,-26,-2,-6,-23,-7,-7,-9,2,-1,-8,-15,-11,-8,-17,3,-4; MA /M: SY='E'; M=-11,0,-2,8,-28,31,15,-17,2,-18,-10,1,-4,-31,-13,-5,-10,-24,-11,-23; MA /M: SY='N'; M=-4,2,18,7,-25,13,7,-1,-1,-24,-26,7,-14,-28,-13,4,-5,-28,-17,-25; MA /M: SY='G'; M=-5,-8,3,-6,-28,-11,-10,25,-14,-27,-20,-4,-13,-24,-13,-5,-14,-23,-21,-23; MA /M: SY='K'; M=-4,8,7,-1,-22,5,3,-9,-3,-22,-23,10,-12,-22,-13,8,1,-29,-13,-17; MA /M: SY='M'; M=0,-18,-22,-29,-6,-16,-22,-24,-18,15,16,-19,21,-2,-24,-14,-4,-25,-7,17; MA /M: SY='T'; M=-3,-7,-11,-18,-3,-12,-13,-21,-19,-5,-6,-6,-4,-8,-20,-1,4,-27,-10,2; MA /M: SY='V'; M=8,-17,-17,-23,-14,-18,-14,-20,-20,5,4,-16,2,3,-20,-4,3,-21,-7,13; MA /M: SY='T'; M=2,-6,-8,-14,-17,-8,-10,-18,-14,-7,-6,-3,-3,-6,-15,3,14,-20,-2,-1; MA /M: SY='N'; M=3,-13,5,-13,-11,-13,-13,-13,-11,-8,-10,-12,-5,0,-14,0,0,-24,-10,-6; MA /M: SY='H'; M=-10,0,-4,-5,-28,6,6,-21,18,-16,-15,3,-6,-12,-15,-6,-9,-14,14,-17; MA /M: SY='T'; M=-7,-6,-3,-11,-23,-1,-9,-3,0,-18,-13,-10,-7,-9,-18,2,4,-17,0,-15; MA /M: SY='V'; M=-7,-2,-18,-23,-20,-15,-17,-26,-20,11,5,-11,4,-6,-17,-9,-4,-26,-9,16; MA /M: SY='V'; M=-6,-2,-13,-18,-19,-8,-15,-24,-16,9,0,-8,8,-10,-21,-5,0,-27,-9,15; MA /I: MI=0; I=-1; MD=0; /M: SY='X'; M=0; D=-1; MA /M: SY='D'; M=1,0,5,15,-24,13,10,-11,-5,-26,-24,9,-15,-32,-10,5,-5,-28,-16,-20; MA /M: SY='T'; M=1,-10,1,-9,-10,-9,-9,-18,-19,-11,-12,-10,-11,-11,-10,22,46,-31,-11,-1; MA /M: SY='D'; M=-19,-9,25,64,-29,0,18,-9,1,-38,-30,0,-29,-38,-11,1,-9,-40,-20,-30; MA /M: SY='Y'; M=-19,-9,-14,-17,-29,-9,-18,-28,19,-2,-2,-9,-2,26,-29,-18,-9,25,72,-12; MA /M: SY='D'; M=-15,0,10,38,-29,9,15,-15,-2,-30,-23,9,-18,-34,-11,-3,-8,-32,-15,-25; MA /M: SY='N'; M=-6,1,26,5,-18,8,1,-10,-1,-18,-22,-1,-13,-21,-15,13,14,-33,-15,-19; MA /M: SY='Y'; M=-20,-12,-20,-24,-28,-16,-22,-30,12,0,2,-14,0,40,-30,-20,-10,26,70,-8; MA /M: SY='A'; M=24,-19,-11,-21,-14,-12,-14,-12,-21,2,-1,-15,-3,-12,-14,5,5,-23,-14,6; MA /M: SY='I'; M=-8,-21,-24,-33,-16,-19,-25,-31,-21,27,24,-24,24,6,-25,-20,-9,-21,-1,20; MA /M: SY='M'; M=-2,-16,-11,-19,-21,-12,-10,-16,-13,5,2,-15,6,-4,-19,-8,-5,-22,-7,5; MA /M: SY='Y'; M=-20,-8,-9,-11,-30,-7,-13,-25,38,-13,-8,-12,-3,14,-26,-16,-14,9,47,-17; MA /M: SY='S'; M=5,-14,-3,-13,-15,-8,-10,-13,-12,-4,-4,-15,-6,-8,-17,10,10,-26,-6,-2; MA /M: SY='C'; M=-9,-24,-16,-29,48,-22,-25,-31,-28,-2,-6,-23,-5,-13,-28,-10,-3,-36,-17,4; MA /M: SY='R'; M=-10,20,13,-1,-24,1,-2,-10,-6,-21,-22,14,-12,-21,-17,0,-1,-27,-14,-16; MA /M: SY='T'; M=-8,-4,-10,-15,-18,-2,-9,-24,-9,-4,1,-6,0,-3,-19,-5,7,-17,4,-4; MA /I: MI=0; I=-1; MD=0; /M: SY='X'; M=0; D=-1; MA /M: SY='H'; M=-3,-5,2,-2,-20,-3,-5,-16,19,-15,-13,-7,-6,-17,-17,1,2,-30,-4,-9; MA /M: SY='F'; M=-10,-12,0,2,-20,-13,-5,-18,-6,-11,-9,-13,-8,3,-19,0,-1,-25,-4,-6; MA /M: SY='D'; M=-9,0,4,11,-26,7,11,-10,-1,-21,-20,-1,-13,-26,-12,3,-4,-29,-14,-19; MA /M: SY='Y'; M=-15,-14,-16,-25,-25,-14,-19,-28,10,6,11,-19,11,18,-25,-18,-9,-5,27,-1; MA /M: SY='A'; M=12,-15,-11,-20,-13,-9,-14,-12,-14,5,-3,-13,11,-10,-17,6,3,-27,-12,10; MA /M: SY='W'; M=-14,-10,-20,-27,-31,-14,-19,-17,-19,-14,-11,-13,-10,13,-24,-18,-12,45,12,-15; MA /M: SY='I'; M=-8,-24,-26,-34,-22,-22,-26,-34,-26,34,29,-28,18,4,-26,-21,-8,-22,-2,27; MA /M: SY='F'; M=-16,-17,-24,-31,-23,-25,-24,-30,-9,7,21,-24,7,41,-30,-24,-10,5,33,1; MA /M: SY='G'; M=10,-15,2,-7,-17,-9,-9,22,-15,-23,-23,-15,-17,-22,-15,19,2,-30,-22,-15; MA /M: SY='R'; M=-19,66,0,-9,-30,10,1,-20,-1,-30,-21,32,-10,-21,-19,-10,-10,-20,-10,-20; MA /M: SY='N'; M=-3,0,17,11,-19,7,4,-8,2,-23,-25,0,-16,-24,-13,15,6,-34,-15,-19; // ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
In order to demonstrate the use of the SBIN programs, we'll have a look
at a set of proteins which all fold into a beta-sandwich, namely:
- 1CEL = T. reesei cellobiohydrolase I
- 1EG1 = T. reesei endoglucanase I
- 1AYH = a hybrid beta-glucanase
- 1LTE = coral tree lectin
The first two of these proteins are very related (which means that their sequence weights should be smaller than those for the other two), the other two are unrelated. The only thing all have in common is the fold class (beta-sandwich).
We shall use 1CEL (chain A) as our "reference molecule", to which the other three will be aligned. This can be done with your favourite superpositioning program, for instance LSQMAN. LSQMAN contains a "BRute_force" command which will automatically align just about any structure with just about any other structure.
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- LSQMAN > read m1 /nfs/pdb/full/1cel.pdb Cell : ( 84.000 86.200 111.800 90.000 90.000 90.000) Old chain |A| becomes chain A Old chain |B| becomes chain B Old chain | | becomes chain C Nr of lines read from file : ( 7524) Nr of atoms in molecule : ( 7038) Nr of chains or models : ( 3) Stripped hydrogen atoms : ( 0) Nr of HETATMs : ( 598) LSQMAN > read m2 /nfs/pdb/full/1ayh.pdb Cell : ( 64.320 78.520 39.300 90.000 90.000 90.000) Old chain | | becomes chain A Nr of lines read from file : ( 1995) Nr of atoms in molecule : ( 1856) Nr of chains or models : ( 1) Stripped hydrogen atoms : ( 0) Nr of HETATMs : ( 159) LSQMAN > br Mol 1 ? (M2) m1 Chain 1 ? (A) Mol 2 ? (M1) m2 Chain 2 ? (A) Fragment length ? ( 50) Fragment step ? ( 25) Min nr of residues to match ? ( 100) Brute-force fit of M1 A And M2 A Atom types | CA | B-factor range used -1000.00 - 10000.00 A2 Fragment length 50 Fragment step size 25 Min matched residues 100 Mol 1 zone to try : (A1-764) Mol 2 zone to try : (A1-373)Try zone : (A1-50) Max match so far : ( 6) RMSD (A) : ( 0.928) Max match so far : ( 10) RMSD (A) : ( 1.600) Max match so far : ( 17) RMSD (A) : ( 2.083) Max match so far : ( 21) RMSD (A) : ( 1.698) Try zone : (A26-75) Max match so far : ( 22) RMSD (A) : ( 1.970) Try zone : (A51-100) Try zone : (A76-125) Max match so far : ( 123) RMSD (A) : ( 1.715)
Max match : ( 123) RMSD (A) : ( 1.715) Mol 1 res : ( 76) Mol 2 res : ( 32)
Regenerating best alignment ... The 123 atoms have an RMS distance of 1.715 A SI = RMS * Nmin / Nmatch = 2.98401 MI = (1+Nmatch)/{(1+W*RMS)*(1+Nmin)} = 0.21242 MC = Maiorov-Crippen RHO (0-2) = 0.12160 RMS delta B for matched atoms = 4.972 A2 Corr. coefficient matched atom Bs = 0.488 Rotation : 0.00118718 0.78486478 0.61966598 -0.99711239 0.04798083 -0.05886189 -0.07593071 -0.61780673 0.78265536 Translation : 47.5729 52.6618 47.1068 CPU total/user/sys : 11.2 11.2 0.0 LSQMAN > apply m1 m2 Bring Mol 2 on top of Mol 1 ... Molecule 1 : (M1) Molecule 2 : (M2) Apply to mol 2 chain : (*) Nr of atoms moved : ( 1856) Resetting ALL operators of mol 2 ... LSQMAN > wr m1 1cel.pdb a Write mol : (M1) Chain id : (A) PDB file : (1cel.pdb) Number of atoms written : ( 3518) LSQMAN > wr m2 1ayh.pdb a Write mol : (M2) Chain id : (A) PDB file : (1ayh.pdb) Number of atoms written : ( 1856) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
When you are done, put all superimposed models in one big PDB file (only CA atoms are really needed). Make sure that the chain names are different for subsequent molecules !!! This can all be done easily with MOLEMAN2:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
MOLEMAN2 > read 1cel.pdb
MOLEMAN2 > append 1eg1.pdb
MOLEMAN2 > append 1ayh.pdb
MOLEMAN2 > append 1lte.pdb
MOLEMAN2 > chain auto
MOLEMAN2 > pdb remark "ALIGNED LSQMAN: 1CEL 1EG1 1AYH 1LTE"
Add REMARK record : (ALIGNED LSQMAN: 1CEL 1EG1 1AYH 1LTE)
2: REMARK ALIGNED LSQMAN: 1CEL 1EG1 1AYH 1LTE
MOLEMAN2 > write aligned.pdb pdb calpha
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Let's first use STRUPAT to see if the sequences of the structurally conserved parts of these four proteins yield any patterns, using the following parameters:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Equivalent CA distance ? ( 3.500) 6 Equivalent CA distance : ( 6.000)Extension CA distance ? ( 6.000) 8 Extension CA distance : ( 8.000)
Try to correct frame-shifts (Y/N) ? (Y) Try to correct frame-shifts (Y/N) : (Y)
Min pattern length ? ( 10) 5 Min pattern length : ( 5) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
There is only one pattern that fits the bill, but it's not a very selective one:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- New structurally conserved stretch ! Starts at residue GLY - 139 molecule 2 @ GLY - 140 molecule 3 @ GLY - 86 molecule 4 @ ALA - 88 |GGGA| ==> [AG] - |LEID| ==> X - |NNVG| ==> X - |GGSL| ==> X - |ASSV| ==> X - |LLFF| ==> X - |YYFF| ==> [FY] - |FLTM| ==> X - |VSYG| ==> X - |SQTP| ==> X - |MMGT| ==> X - Nr of residues : ( 11) RMS (RMSD) (A) : ( 1.407) PROSITE pattern : ([AG] - X(5) - [FY]) Length : ( 7) Score 1 : ( 1.386) Score 2 : ( 13) Nr of matches to random sequence : ( 25298) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Clearly, the sequence similarities (if any) between these proteins are too low to yield any useful patterns. Let's use STRUPRO then to generate a profile with the following input:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Library file with matrix ? (/nfs/public/lib/strupro_blosum45.lib) ../strupro_blosum45.lib Library file with matrix : (../strupro_blosum45.lib) Comment : (! BLOSUM 45 matrix made from BLOCKS v. 5.0 and scaled in half- bits.) Comment : (! ARNDCQEGHILKMFPSTWYVBZX) Comment : (! integer matrix)Equivalent CA distance ? ( 5.000) 6 Equivalent CA distance : ( 6.000)
Extension CA distance ? ( 8.000) Extension CA distance : ( 8.000)
Try to correct frame-shifts (Y/N) ? (Y) Try to correct frame-shifts (Y/N) : (Y)
Min fragment length ? ( 5) Min fragment length : ( 5)
Sequences may be weighted: U = uniform weights R = rms(rmsd) weights S = sequence distance weights Weighting scheme ? (S) Weighting scheme : (S) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
The result:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Nr of residues in profile : ( 189)Sequence identity for these residues only: % Seq id mol # 1 -> 100.0 49.7 11.1 7.9 % Seq id mol # 2 -> 49.7 100.0 10.1 7.9 % Seq id mol # 3 -> 11.1 10.1 100.0 9.0 % Seq id mol # 4 -> 7.9 7.9 9.0 100.0
Average sequence identity (%) : ( 15.961) St. dev. : ( 15.146) Minimum : ( 7.937) Maximum : ( 49.735)
Sum of maximum random scores : ( 3275) Sum AVE+3SIGMA random scores : ( 1088)
Score for molecule 1 = 3780 Score for molecule 2 = 3806 Score for molecule 3 = 3205 Score for molecule 4 = 2879
Minimum raw score : ( 2400) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
The profile looks as follows:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ID STRUPRO; MATRIX. AC PS99999; DT JAN-1900 (CREATED); DE Created by STRUPRO V. 971103/1.0 at Mon Nov 3 17:01:47 1997 for user gerard CC CC Substitution matrix file : ../strupro_blosum45.lib CC Nr of structures used : 4 CC Equivalent CA distance (A) : 6.000000 CC Extension CA distance (A) : 8.000000 CC Frameshift correction used CC Min fragment length : 5 CC Weighting scheme : S CC MA /GENERAL_SPEC: ALPHABET='ARNDCEQGHILKMFPSTWYV'; LENGTH=189; MA TOPOLOGY=LINEAR; MA /DISJOINT: DEFINITION=PROTECT; N1=1; N2=189; MA /CUT_OFF: LEVEL=0; SCORE= 2400; MA /DEFAULT: MI=-100; I=-10; IM=0 ; MD=-100; D=-3; DM=0; MA /M: SY='T'; M=-2,-15,-13,-19,-18,-15,-17,-17,-22,-4,-12,-15,-10,-6,-19,5,13,5,-4,4; MA /M: SY='Y'; M=-15,0,-11,-12,-27,-7,2,-25,-2,-14,-9,4,-7,15,-19,-13,-10,0,23,-14; MA /M: SY='K'; M=-13,12,-5,0,-30,8,15,-23,0,-22,-20,24,-10,-15,-12,-10,-10,-10,10,-20; MA /M: SY='C'; M=8,-22,-12,-22,53,-20,-20,-20,-25,-20,-15,-20,-15,-17,-25,3,8,-37,-22,-5; MA /M: SY='S'; M=7,-10,7,-3,-10,-3,-3,-6,-13,-17,-24,-10,-17,-17,-10,34,28,-37,-17,-7; MA /I: MI=0; I=-1; MD=0; /M: SY='X'; M=0; D=-1;[...]
MA /I: MI=0; I=-1; MD=0; /M: SY='X'; M=0; D=-1; MA /M: SY='N'; M=6,11,17,-1,-18,0,-2,-5,-4,-20,-23,3,-15,-20,-15,12,2,-30,-17,-16; MA /M: SY='D'; M=-10,-8,19,24,-28,-1,14,12,-3,-33,-28,-3,-23,-30,-13,2,-11,-32,-23,-30; MA /M: SY='A'; M=27,-13,11,-6,-12,-5,-5,0,-11,-15,-19,-8,-15,-20,-12,17,5,-29,-20,-9; MA /M: SY='P'; M=5,-14,5,-4,-27,-7,-3,-9,-12,-17,-25,-7,-17,-25,34,1,-5,-30,-25,-22; MA /M: SY='N'; M=-10,-3,22,7,-28,12,5,-11,2,-20,-27,0,-15,-28,13,2,-5,-32,-20,-30; MA /M: SY='T'; M=9,-15,-11,-18,-13,-13,-13,-18,-20,-2,7,-16,-2,-7,-16,4,21,-25,-10,3; MA /M: SY='H'; M=-18,4,-1,-6,-30,4,-3,-23,51,-22,-17,4,-2,-8,-20,-13,-15,-11,30,-22; MA /M: SY='V'; M=9,-17,-12,-2,-15,-17,-12,-17,-20,1,-6,-12,-6,-16,-19,-2,-3,-30,-15,16; MA /M: SY='V'; M=-3,-9,-14,-17,-16,0,-11,-25,-16,6,-3,-9,2,-14,-20,0,9,-27,-10,15; MA /M: SY='Y'; M=-20,-15,-20,-29,-25,-24,-25,-30,1,0,5,-19,0,53,-30,-20,-10,21,57,-5; MA /M: SY='S'; M=2,-10,13,20,-16,0,6,-3,-7,-26,-30,-7,-23,-26,-10,29,12,-40,-20,-16; MA /M: SY='W'; M=-15,-11,7,-12,-36,-11,-16,-11,-11,-20,-25,-11,-20,-4,-25,-17,-16,61,7,-30; MA /M: SY='I'; M=-2,-22,-15,-27,-19,-18,-22,-27,-25,27,5,-22,7,-5,-20,-2,0,-28,-8,26; MA /M: SY='R'; M=-15,27,-5,-12,-27,-3,-2,-23,-10,-22,-18,25,-7,0,-17,-13,-10,-12,0,-15; MA /M: SY='W'; M=-17,-12,-20,-19,-32,-12,-5,-25,-6,-12,-7,-12,-10,22,-22,-20,-15,38,32,-17; MA /M: SY='G'; M=13,-17,-3,-13,-19,-15,-15,27,-20,-24,-19,-15,-15,-22,-15,8,5,-23,-22,-14; MA /M: SY='S'; M=-2,-12,5,14,-22,-2,5,-7,-10,-25,-30,-8,-22,-27,14,19,6,-38,-22,-19; // ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
The program also generates a structure-based sequence alignment:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ! ! Sequence alignment file ! Created by STRUPRO V. 971103/1.0 at Mon Nov 3 16:59:34 1997 for user gerard ! ! REMARK Created by MOLEMAN2 V. 970924/2.1.3 at Mon Nov 3 16:48:35 1997 for user gerard ! REMARK ALIGNED LSQMAN: 1CEL 1EG1 1AYH 1LTE! NOT ALIGNED MOL 1 FROM PCA- 1 TO THR- 15 ! ASACTLQSETHPPLT ! NOT ALIGNED MOL 2 FROM PYR- 1 TO THR- 15 ! AQPGTSTPEVHPKLT ! NOT ALIGNED MOL 3 FROM GLN- 1 TO TRP- 208 ! QTGGSFFEPFNSYNSGTWEKADGYSNGGVFNCTWRANNVNFTNDGKLKLGLTSSAYNKFDCAEYRSTNIYGYGLYEVSMKPAKNTGIVSSFFTYTGPAHGTQWDEIDIEFLGKDTTKVQFNYYTNG ! NOT ALIGNED MOL 4 FROM VAL- 1 TO TRP- 231 ! VETISFSFSEFEPGNDNLTLQGASLITQSGVLQLTKINQNGMPAWDSTGRTLYAKPVHIWDMTTGTVASFETRFSFSIEQPYTRPLPADGLVFFMGPTKSKPAQGYGYLGIFNQSKQDNSYQTLGV ! ALIGNED MOL 1 FROM TRP- 16 TO SER- 20 WQKCS- ! ALIGNED MOL 2 FROM THR- 16 TO THR- 20 TYKCT- ! ALIGNED MOL 3 FROM VAL- 209 TO SER- 213 VKYTS- ! ALIGNED MOL 4 FROM SER- 232 TO SER- 236 SFQAS-
! NOT ALIGNED MOL 1 FROM SER- 21 TO GLN- 28 ! SGGTCTQQ ! NOT ALIGNED MOL 2 FROM LYS- 21 TO GLN- 28 ! KSGGCVAQ ! ALIGNED MOL 1 FROM THR- 29 TO ASP- 35 TGSVVID- ! ALIGNED MOL 2 FROM ASP- 29 TO ASP- 35 DTSVVLD- ! ALIGNED MOL 3 FROM GLY- 16 TO ASP- 22 GTWEKAD- ! ALIGNED MOL 4 FROM ASP- 16 TO GLY- 22 DNLTLQG-
[...]
! NOT ALIGNED MOL 1 FROM ILE- 426 TO GLY- 434 ! IGSTGNPSG ! NOT ALIGNED MOL 2 FROM ILE- 367 TO THR- 371 ! IGSTT ! NOT ALIGNED MOL 3 FROM ASN- 214 TO ASN- 214 ! N ! NOT ALIGNED MOL 4 FROM LEU- 237 TO GLU- 239 ! LPE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Let's scan our profile against SWISS-PROT (release 34) using the "pfsearch" program ("pftools" suite of programs). Using the "-a" flag will print a score for every sequence in the database; this will enable us to calculate the average score etc. and, from that, Z-scores.
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ../pftools/pfsearch -a aligned.prf /nfs/scr_uu5/gerard/sprot34.dat > pfsearch_all.log ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Sort the output file so that the top-scoring entries are at the top:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- sort -nr pfsearch_all.log > pfsearch_all.sorted ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Let's see what the top-scoring entries are:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- % 1004 gerard sarek 16:03:00 prosite/1cel > head -20 pfsearch_all.sorted 3662 P07981|GUN1_TRIRE ENDOGLUCANASE EG-1 PRECURSOR (EC 3.2.1.4) (ENDO-1,4-BETA-GLUCANASE) (CELLULASE). 3577 P00725|GUX1_TRIRE EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDROLASE I) (1,4-BETA-CELLOBIOHYDROLASE). 3557 P19355|GUX1_TRIVI EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDROLASE) (1,4- BETA-CELLOBIOHYDROLASE). 2937 P13860|GUX1_PHACH EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDROLASE I) (1,4-BETA-CELLOBIOHYDROLASE). 2876 P15828|GUX1_HUMGR EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDROLASE I) (1,4- BETA-CELLOBIOHYDROLASE) (BETA-GLU 2836 P23904|GUB_BACMA BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA-D-GLUCAN 4-GLUCANOHYD 2814 P38676|GUX1_NEUCR EXOGLUCANASE 1 PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDROLASE 1) (1,4-BETA-CELLOBIOHYDROLASE). 2805 Q06886|GUX1_PENJA EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDROLASE I) (1,4-BETA-CELLOBIOHYDROLASE). 2741 P45797|GUB_BACPO BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA-D-GLUCAN 4-GLUCANOHYD 2692 P46238|GUXC_FUSOX PUTATIVE EXOGLUCANASE TYPE C PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDROLASE I) (1,4-BETA-CELLOBIOHYDROLA 2492 P46237|GUNC_FUSOX PUTATIVE ENDOGLUCANASE TYPE C PRECURSOR (EC 3.2.1.4) (ENDO-1,4-BETA- GLUCANASE) (CELLULASE). 2334 P16404|LEC_ERYCO LECTIN PRECURSOR. 2326 P04957|GUB_BACSU BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA-D-GLUCAN 4-GLUCANOHYD 2317 P07980|GUB_BACAM BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA-D-GLUCAN 4-GLUCANOHYD 2194 P27051|GUB_BACLI BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA-D-GLUCAN 4-GLUCANOHYD 1971 P37073|GUB_BACBR BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA-D-GLUCAN 4-GLUCANOHYD 1863 P29716|GUB_CLOTM BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA-D-GLUCAN 4-GLUCANOHYD 1400 Q01806|LEC1_MEDTR LECTIN 1 PRECURSOR. 1364 P02867|LEC_PEA LECTIN PRECURSOR. 1354 P29257|LEC2_CYTSC 2-ACETAMIDO-2-DEOXY-D-GALACTOSE-BINDING SEED LECTIN II (CSII). ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
As hoped and expected, these are all CBH I, EG I, beta-glucanase and lectin-type proteins.
Let's put the (arbitrary) scores on a better scale. One option is to use the "pfscale" program ("pftools" package):
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ../pftools/pfscale pfsearch_all.sorted > pfsearch.scale ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
This gives us:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
# -LogP = -2.5931 + 0.00803501 * raw-score
#
# rank raw-score -logFreq -logProb
#
1 3662.00 7.4517 26.8312
2 3577.00 6.9746 26.1482
3 3557.00 6.7527 25.9875
4 2937.00 6.6066 21.0058
5 2876.00 6.4975 20.5157
6 2836.00 6.4103 20.1942
7 2814.00 6.3378 20.0175
8 2805.00 6.2756 19.9452
9 2741.00 6.2213 19.4309
10 2692.00 6.1730 19.0372
11 2492.00 6.1295 17.4302
12 2334.00 6.0900 16.1607
13 2326.00 6.0538 16.0964
14 2317.00 6.0203 16.0241
15 2194.00 5.9893 15.0358
16 1971.00 5.9603 13.2440
17 1863.00 5.9332 12.3762
18 1400.00 5.9076 8.6560
19 1364.00 5.8835 8.3667
20 1354.00 5.8606 8.2864
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
The expression for -LogP can be inserted into the profile if you wish.
Alternatively, simple Z-scores can be calculated, e.g. with ZPROF:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ZPROF < pfsearch_all.sorted > zprof.top ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
This gives:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Working ...Nr of sequences scored : ( 59021) Average : ( 527.689) St.dev. : ( 186.136) Minimum : ( 47.000) Maximum : (3662.000)
Remove "outliers" and re-calc ... Nr of sequences left : ( 58998) Average : ( 526.978) St.dev. : ( 182.065) Minimum : ( 47.000) Maximum : (1231.000)
Remove "outliers" and re-calc ... Nr of sequences left : ( 58998) Average : ( 526.978) St.dev. : ( 182.065) Minimum : ( 47.000) Maximum : (1231.000)
Converged !
MA /NORMALIZATION: MODE=1; FUNCTION=LINEAR; MA R1= -2.89445400; R2= 0.00549255; TEXT ='Z-score'; MA /CUT_OFF: LEVEL=0; SCORE= 1255; N_SCORE= 4.00000000; MODE=1;
Z-score of 0 requires raw score 527 Z-score of 1 requires raw score 709 Z-score of 2 requires raw score 891 Z-score of 3 requires raw score 1073 Z-score of 4 requires raw score 1255 Z-score of 5 requires raw score 1437 Z-score of 6 requires raw score 1619 Z-score of 7 requires raw score 1801 Z-score of 8 requires raw score 1983 Z-score of 9 requires raw score 2166 Z-score of 10 requires raw score 2348 Z-score of 11 requires raw score 2530 Z-score of 12 requires raw score 2712 Z-score of 13 requires raw score 2894 Z-score of 14 requires raw score 3076 Z-score of 15 requires raw score 3258 Z-score of 16 requires raw score 3440 Z-score of 17 requires raw score 3622 Z-score of 18 requires raw score 3804 Z-score of 19 requires raw score 3986 Z-score of 20 requires raw score 4168 Z-score of 21 requires raw score 4350 Z-score of 22 requires raw score 4532 Z-score of 23 requires raw score 4714 Z-score of 24 requires raw score 4897 Z-score of 25 requires raw score 5079
1 17.22 3662 P07981|GUN1_TRIRE ENDOGLUCANASE EG-1 PRECURSOR (EC 3.2.1.4) (ENDO-1,4-BET 2 16.75 3577 P00725|GUX1_TRIRE EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 3 16.64 3557 P19355|GUX1_TRIVI EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 4 13.24 2937 P13860|GUX1_PHACH EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 5 12.90 2876 P15828|GUX1_HUMGR EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 6 12.68 2836 P23904|GUB_BACMA BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 7 12.56 2814 P38676|GUX1_NEUCR EXOGLUCANASE 1 PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 8 12.51 2805 Q06886|GUX1_PENJA EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 9 12.16 2741 P45797|GUB_BACPO BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 10 11.89 2692 P46238|GUXC_FUSOX PUTATIVE EXOGLUCANASE TYPE C PRECURSOR (EC 3.2.1.91) (E 11 10.79 2492 P46237|GUNC_FUSOX PUTATIVE ENDOGLUCANASE TYPE C PRECURSOR (EC 3.2.1.4) (E 12 9.93 2334 P16404|LEC_ERYCO LECTIN PRECURSOR. 13 9.88 2326 P04957|GUB_BACSU BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 14 9.83 2317 P07980|GUB_BACAM BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 15 9.16 2194 P27051|GUB_BACLI BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 16 7.93 1971 P37073|GUB_BACBR BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 17 7.34 1863 P29716|GUB_CLOTM BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 18 4.80 1400 Q01806|LEC1_MEDTR LECTIN 1 PRECURSOR. 19 4.60 1364 P02867|LEC_PEA LECTIN PRECURSOR. 20 4.54 1354 P29257|LEC2_CYTSC 2-ACETAMIDO-2-DEOXY-D-GALACTOSE-BINDING SEED LECTIN II 21 4.45 1338 P53301|YG46_YEAST HYPOTHETICAL 52.8 KD PROTEIN IN BUB1-HIP1 INTERGENIC RE 22 4.41 1329 P05046|LEC_SOYBN LECTIN PRECURSOR. 23 4.34 1317 Q01807|LEC2_MEDTR TRUNCATED LECTIN 2 PRECURSOR. 24 3.87 1231 P33693|EXOK_RHIME SUCCINOGLYCAN BIOSYNTHESIS PROTEIN EXOK.
Z-score cut-off : ( 4.000) Nr of "hits" : ( 23) % of database : ( 0.039) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
There is one protein in this list of those that have Z-score > 4 which is not an obvious hit: YG46_YEAST HYPOTHETICAL 52.8 KD PROTEIN.
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
ID YG46_YEAST STANDARD; PRT; 507 AA.
AC P53301;
DT 01-OCT-1996 (REL. 34, CREATED)
DT 01-OCT-1996 (REL. 34, LAST SEQUENCE UPDATE)
DT 01-FEB-1997 (REL. 35, LAST ANNOTATION UPDATE)
DE HYPOTHETICAL 52.8 KD PROTEIN IN BUB1-HIP1 INTERGENIC REGION.
GN YGR189C OR G7553.
OS SACCHAROMYCES CEREVISIAE (BAKER'S YEAST).
OC EUKARYOTA; FUNGI; ASCOMYCOTINA; HEMIASCOMYCETES.
RN [1]
RP SEQUENCE FROM N.A.
RC STRAIN=S288C;
RX MEDLINE; 97279231. [NCBI, Geneva]
RA ARROYO J., GARCIA-GONZALEZ M., GARCIA-SAEZ M.I., SANCHEZ-PEREZ M.,
RA NOMBELA C.;
RL YEAST 13:357-363(1997).
CC -!- SIMILARITY: SOME, TO YEAST UTR2.
DR EMBL; Z72974; E243566; -. [EMBL / GenBank / DDBJ] [CoDingSequence]
DR EMBL; X99074; E252633; -. [EMBL / GenBank / DDBJ] [CoDingSequence]
DR PRODOM [Domain structure / List of seq. sharing at least 1 domain]
DR SWISS-2DPAGE; GET REGION ON 2D PAGE.
KW HYPOTHETICAL PROTEIN.
FT DOMAIN 63 66 POLY-SER.
FT DOMAIN 301 310 POLY-SER.
FT DOMAIN 345 357 POLY-SER.
FT DOMAIN 387 391 POLY-SER.
FT DOMAIN 467 470 POLY-SER.
SQ SEQUENCE 507 AA; 52757 MW; 4CC1838E CRC32;
MKVLDLLTVL SASSLLSTFA AAESTATADS TTAASSTASC NPLKTTGCTP DTALATSFSE
DFSSSSKWFT DLKHAGEIKY GSDGLSMTLA KRYDNPSLKS NFYIMYGKLE VILKAANGTG
IVSSFYLQSD DLDEIDIEWV GGDNTQFQSN FFSKGDTTTY DRGEFHGVDT PTDKFHNYTL
DWAMDKTTWY LDGESVRVLS NTSSEGYPQS PMYLMMGIWA GGDPDNAAGT IEWAGGETNY
NDAPFTMYIE KVIVTDYSTG KKYTYGDQSG SWESIEADGG SIYGRYDQAQ EDFAVLANGG
SISSSSTSSS TVSSSASSTV SSSVSSTVSS SASSTVSSSV SSTVSSSSSV SSSSSTSPSS
STATSSKTLA SSSVTTSSSI SSFEKQSSSS SKKTVASSST SESIISSTKT PATVSSTTRS
TVAPTTQQSS VSSDSPVQDK GGVATSSNDV TSSTTQISSK YTSTIQSSSS EASSTNSVQI
SNGADLAQSL PREGKLFSVL VALLALL
//
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
The similarity of this sequence to those used to construct the profile is not noted in the entry. However, the link to the PRODOM database reveals similarities in particular to several beta-glucanases. The entry EXOK_RHIME also shows similarities, and this is also the first protein to have a Z-score below 4.0. Let's see what some of the other proteins with Z-score < 4.0 are:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- 1231 P33693|EXOK_RHIME SUCCINOGLYCAN BIOSYNTHESIS PROTEIN EXOK. 1229 P02871|LEC_VICFA FAVIN (LECTIN). 1219 P17989|GUB_FIBSU BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA-D-GLUCAN 4-GLUCANOHYD 1199 P22972|LEC1_ULEEU ANTI-H(O) LECTIN I (UEA-I). 1198 P16030|LEC_BAUPU LECTIN PRECURSOR. 1183 P23558|LEC1_LABAL LECTIN I (SEED LECTIN ANTI-H(O)) (LAA-I). 1167 P22973|LEC2_ULEEU ANTI-H(O) LECTIN II (UEA-II). 1131 P35694|BRU1_SOYBN BRASSINOSTEROID-REGULATED PROTEIN BRU1. 1125 P15231|PHAM_PHAVU LEUCOAGGLUTINATING PHYTOHEMAGGLUTININ PRECURSOR (PHA-L). 1123 P16349|LEC_LATSP LECTIN. 1115 P05088|PHAE_PHAVU ERYTHROAGGLUTINATING PHYTOHEMAGGLUTININ PRECURSOR (PHA-E). 1115 P02873|LEC_PHAVU LECTIN PRECURSOR (ALPHA-AMYLASE INHIBITOR). 1111 P39795|TREC_BACSU TREHALOSE-6-PHOSPHATE HYDROLASE (EC 3.2.1.93) (ALPHA,ALPHA- PHOSPHOTREHALASE). 1110 P02874|LEC_ONOVI LECTIN. 1102 P05087|PHAL_PHAVU LEUCOAGGLUTINATING PHYTOHEMAGGLUTININ PRECURSOR (PHA-L). 1082 P24806|MER5_ARATH MERI-5 PROTEIN. 1074 P45798|GUB_RHOMR BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA-D-GLUCAN 4-GLUCANOHYD 1066 P02870|LEC_LENCU LECTIN. 1065 P02866|CONA_CANEN CONCANAVALIN A PRECURSOR (CON A). 1034 P19588|LEC5_DOLBI LECTIN DB58 PRECURSOR. 1031 P14894|CONA_CANGL CONCANAVALIN A PRECURSOR (CON A). 1030 P24146|LEC4_GRISI LECTIN IV (GS4). 1029 P43478|CGKA_ALTCA KAPPA-CARRAGEENASE PRECURSOR (EC 3.2.1.83). 1027 P32623|UTR2_YEAST UTR2 PROTEIN (UNKNOWN TRANSCRIPT 2 PROTEIN). 1024 P05045|LEC1_DOLBI SEED LECTIN SUBUNITS I AND II PRECURSOR. 1024 P04122|LECB_LATOC LECTIN BETA-1 AND BETA-2 CHAINS. 1019 P07067|VG37_BPT2 TAIL FIBER PROTEIN GP37. 1018 P20847|GUN1_BUTFI ENDOGLUCANASE 1 (EC 3.2.1.4) (ENDO-1,4-BETA-GLUCANASE) (CELLULASE). 1002 P32818|AMYM_BACAD MALTOGENIC ALPHA-AMYLASE PRECURSOR (EC 3.2.1.133) (GLUCAN 1,4-ALPHA- MALTOHYDROLASE). ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
In fact, many of these sequences are "hits": more beta-glucanases and lectins, and concanavalins. It therefore appears not impossible that some of the "unknown" hits may also adopt a beta-sandwich fold.
We can try to "refine" the profile by adding more sequences to those that were used to generate the original profile. Let's first include a scoring formula into the profile, based on the output of ZPROF (except that we use a cut-off Z-score of 3.0 instead of 4.0):
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- MA /NORMALIZATION: MODE=1; FUNCTION=LINEAR; MA R1= -2.89445400; R2= 0.00549255; TEXT ='Z-score'; MA /CUT_OFF: LEVEL=0; SCORE= 1073; N_SCORE= 3.00000000; MODE=1; ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Now scan the database again with the original profile, but use the "-ry" flags for the "pfsearch" program, so it will only display hits as well as an alignment of each sequence with the profile:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ../pftools/pfsearch -ry aligned.prf /nfs/scr_uu5/gerard/sprot34.dat | & tee strupro.hits 1131 pos. 28 - 250 P35694|BRU1_SOYBN BRASSINOSTEROID-REGULATED PROTEIN BRU1. # # P 4 CSXDNWTVADXNGVTTS----------------XSLQLGQITQ-XNVCARYYYMNXYKMF -164 # S 28 CA--GSFYQD-----FDltwggdrakifnggqlLSLSLDKVSGsGFKSKKEYLFG----- -209 # # P 47 HLWXGLYSFDVDPAEQP------XGLNGSFFMGPM-XCCDEMDIEFDNXPHIALNPHXCD -111 # S 76 RID----------MQLKlvagnsAGTVTAYYLSSQgPTHDEIDFEFLG----NLSGD--- -166 # # P 100 SGGCEWNPY--XTGPFS-------------XLDTSKFHTVVFQWDPSXKITRYYQ----X -70 # S 119 -------PYilHTNIFTqgkgnreqqfylwFDPTRNFHTYSIIWKPQ-HIIFLVDntpiR -113 # # P 141 TFPQAXNTLTAXGLANMPKAPXSWMDIMMSLWNGTXFSNPWLD----------------- -27 # S 171 VFKNA-EPLGV----PFPKNQ--PMRIYSSLWNAD----DWATrgglvktdwskapftay -65 # # P 183 -----------------XGAPNDAEXNDAPNTHVVYS -7 # S 220 yrnfkaiefsskssisnSGAEYEAN------ELDAYS -34 # [...] ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
The output file can be converted into a mutliple-sequence alignment file with PRF2MSEQ, which can in turn serve as input to MSEQPRO to generate a new profile using all matching sequences:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- PRF2MSEQ < strupro.hits > hits.mseq ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
The multiple-sequence alignment file may look as follows:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
! 1131 pos. 28 - 250 P35694|BRU1_SOYBN BRASSINOSTEROID-REGULATED PROTEIN BRU1.
CA GSFYQD FD SLSLDKVSG FKSKKEYLFG RID MQLK GTVTAYYLSSQ THDEIDFEFLG NLSGD PY TNIFT DPTRNFHTYSIIW
! 1231 pos. 64 - 268 P33693|EXOK_RHIME SUCCINOGLYCAN BIOSYNTHESIS PROTEIN EXOK.
CT WSKKQ VKTV ILELTFEEK FACGEIQTRK RFGYG TYEARIKAADGS GLNSAFFTYIG PHDEIDFEVLG AKVQINQY SAKGGNEFLAD VPGG ANQGFNDYAFVW
! 2317 pos. 40 - 236 P07980|GUB_BACAM BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA
S DGYSNGD NNVSMT EMRLALTSP FDCGENRSVQ TYGYG LYEVRMKPAKNT GIVSSFFTYTG PWDEIDIEFLG TKVQFNYY NGAGNHEKFAD LG DAANAYHTYAFDW
! 1971 pos. 41 - 251 P37073|GUB_BACBR BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1,4 GLUCANASE) (1,3-1,4-BETA
FYES FD AGVWTN RLTIAKKTT SARNYKAG NDFYHYG LFEVSMKPAKVE GTVSSFFTYTG PWDEIDIEFLG TRIQFNYF NGVGGNEFYYD LG DASESFNTYAFEW
[...]
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Now run MSEQPRO with these aligned sequences as input:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Random-number seed ? ( 123456) Random-number seed : ( 123456) => Random number generator initialised with seed : 123456Generating random sequence ... Target composition : ( 0.081 0.044 0.046 0.058 0.019 0.058 0.037 0.080 0.022 0.053 0.081 0.059 0.020 0.040 0.047 0.068 0.063 0.016 0.038 0.071) Working ... Actual composition : ( 0.081 0.044 0.046 0.058 0.019 0.057 0.037 0.080 0.022 0.053 0.081 0.060 0.020 0.040 0.046 0.068 0.063 0.015 0.038 0.070)
Library file with matrix ? (/nfs/public/lib/strupro_blosum45.lib) ../strupro_blosum45.lib Library file with matrix : (../strupro_blosum45.lib) Comment : (! BLOSUM 45 matrix made from BLOCKS v. 5.0 and scaled in half- bits.) Comment : (! ARNDCQEGHILKMFPSTWYVBZX) Comment : (! integer matrix)
Min fragment length ? ( 5) Min fragment length : ( 5)
Sequences may be weighted: U = uniform weights S = sequence distance weights Weighting scheme ? (S) Weighting scheme : (S)
Name of sequence file ? (aligned.seq) hits.mseq Name of sequence file : (hits.mseq)
Name of profile file ? (aligned.prf) hits.prf Name of profile file : (hits.prf)
Remark : (! 1131 pos. 28 - 250 P35694|BRU1_SOYBN BRASSINOSTEROID- REGULATED PROTEIN BRU1.) Remark : (! 1231 pos. 64 - 268 P33693|EXOK_RHIME SUCCINOGLYCAN BIOSYNTHESIS PROTEIN EXOK.)
[...]
Remark : (! 1338 pos. 23 - 287 P53301|YG46_YEAST HYPOTHETICAL 52.8 KD PROTEIN IN BUB1-HIP1 INTERGENIC REGION.) Nr of sequences : ( 40) Nr of residues : ( 209) SEQ > (---CA--GSFYQD-----FD-SLSLDKVSG-FKSKKEYLFG-----RID----------MQLK- GTVTAYYLSSQ-THDEIDFEFLG----NLSGD----------PY-TNIFT-DPTRNFHTYSIIWKPQ- HIIFLVD-VFKNA-EPLGV----PFPKNQ--PMRIYSSLWNAD----DWAT-GAEYEAN------ELDAYS) SEQ > (---CT---WSKKQ---VKTV-ILELTFEEK-FACGEIQTRK---RFGYG--TYEARIKAADGS- GLNSAFFTYIG-PHDEIDFEVLG-AKVQINQY-SAKGGNEFLAD--VPGG--ANQGFNDYAFVWEKN- RIRYYVN-----G-HEVTD--PAKIPVNA---QKIFFSLWGTD-TLTDWMG-GDECQFA-AQS)
[...]
SEQ > (--EST-DSTTAAS-NPLKTT-ALATSFSED-FSSSSKWFTD-AGEIKYG-GKLEVILKAANGT- GIVSSFYLQSD-DLDEIDIEWVG-DNTQFQSN-----------F----FS-TPTDKFHNYTLDWAMD- KTTWYLD-SVRVL-NTSSE------GYPQ-SPMYLMMGIWAGG-GTIEWAG-GSWESIE-ADGGSIYGRYD)
Nr of positions without INDELs : ( 60) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- New conserved stretch ! Length : ( 11) Calculating sequence distances ... Weights converged : ( 110000) Largest shift (%) : ( 0.865) Weights : ( 0.028 0.047 0.007 0.007 0.007 0.013 0.013 0.007 0.007 0.044 0.062 0.048 0.046 0.011 0.011 0.023 0.023 0.011 0.011 0.024 0.023 0.021 0.011 0.019 0.021 0.037 0.022 0.051 0.017 0.032 0.016 0.036 0.028 0.013 0.028 0.016 0.014 0.014 0.087 0.045)AA-TYPE : ALA ARG ASN ASP CYS GLU GLN GLY HIS ILE LEU LYS MET PHE PRO SER THR TRP TYR VAL |TPPPPQQPPPDCCCCCCCCCSSSSSSKSTNTKGTTNNNHD| PROFILE : -7 -11 3 1 6 -7 -4 -13 -4 -25 -25 -9 -18 -24 -5 5 1 -37 -19 -19 |HHWWWWWWWWNCCCCCCCCCNIYNINNYIETGDAWAFFYL| PROFILE : -12 -15 -5 -15 2 -12 -17 -18 -3 -14 -12 -16 -10 -5 -27 -12 -11 0 6 -15 |DDDDDDDDDVGNNSSATSSPQQQQQQQQHTQDQQDHHHDD| PROFILE : -10 -7 11 27 -26 2 19 -8 2 -28 -23 -2 -20 -29 -8 3 -5 -34 -17 -23 |EEEEEEEEEEEEEEEEEEEEITTITIVTTVTTVTETTTSE| PROFILE : -5 -2 -3 -8 -22 25 3 -21 -5 -8 -13 -2 -2 -25 -12 6 7 -25 -10 -10 |IIIIIIIIIVIMLMMMMMMMIVVVVIVLVVVVVVIVVVMI| PROFILE : -7 -21 -24 -33 -20 -18 -27 -31 -22 33 20 -22 27 1 -24 -18 -7 -23 -3 30 |DDDDDDDDDDDDDDDDDDDDAAAAAAAGAAATAADAAATD| PROFILE : 5 -14 7 29 -21 -5 5 -4 -10 -27 -21 -5 -21 -30 -11 5 0 -32 -19 -17 |FFIIIIIIIIIIIIIIIIIIVVVVVVVVVVVVVVFVVVVI| PROFILE : -6 -24 -25 -35 -19 -27 -30 -34 -29 35 14 -25 13 8 -26 -15 -5 -22 -2 36 |EEEEEEEEEEMLWWWWWWWWEEEEEEEEEEEEEEEEEEGE| PROFILE : -11 0 -9 -11 -32 33 4 -13 -1 -17 -15 -1 0 -26 -16 -9 -14 7 -4 -26 |FVFFFFFFFVEEEEEEEEEEFIFFIFFFFFFFFFFFFFEW| PROFILE : -14 -10 -15 -25 -24 -2 -12 -26 -11 -3 -2 -14 1 23 -22 -13 -10 1 11 -6 |LLLLLLLLLLHGAAAAAAAADDDDDDDDDDDDDDLDDDMV| PROFILE : -4 -14 -2 13 -23 -7 -2 -11 -3 -14 -3 -11 -4 -19 -17 -7 -9 -29 -12 -10 |GGGGGGGGGGVNNNNNNNNNTTTTTSTTTTTTTTGTTTSG| PROFILE : -1 -11 12 -4 -17 -10 -10 9 -13 -18 -20 -11 -15 -18 -16 13 15 -31 -18 -12
Random sequence tests : 1999990 Average, St.dev. : -93.2 46.5 Minimum, Maximum : -270.0 149.0 Z-min, Z-max : -3.80 5.20
Mol # 1 Raw score = 182 Z-score = 5.91 Mol # 2 Raw score = 147 Z-score = 5.16 Mol # 3 Raw score = 206 Z-score = 6.43 Mol # 4 Raw score = 206 Z-score = 6.43 Mol # 5 Raw score = 206 Z-score = 6.43 Mol # 6 Raw score = 207 Z-score = 6.45 Mol # 7 Raw score = 207 Z-score = 6.45 Mol # 8 Raw score = 206 Z-score = 6.43 Mol # 9 Raw score = 206 Z-score = 6.43 Mol # 10 Raw score = 124 Z-score = 4.67 Mol # 11 Raw score = 93 Z-score = 4.00 Mol # 12 Raw score = 119 Z-score = 4.56 Mol # 13 Raw score = 141 Z-score = 5.03 Mol # 14 Raw score = 140 Z-score = 5.01 Mol # 15 Raw score = 140 Z-score = 5.01 Mol # 16 Raw score = 127 Z-score = 4.73 Mol # 17 Raw score = 132 Z-score = 4.84 Mol # 18 Raw score = 140 Z-score = 5.01 Mol # 19 Raw score = 140 Z-score = 5.01 Mol # 20 Raw score = 129 Z-score = 4.78 Mol # 21 Raw score = 169 Z-score = 5.63 Mol # 22 Raw score = 146 Z-score = 5.14 Mol # 23 Raw score = 192 Z-score = 6.13 Mol # 24 Raw score = 166 Z-score = 5.57 Mol # 25 Raw score = 146 Z-score = 5.14 Mol # 26 Raw score = 167 Z-score = 5.59 Mol # 27 Raw score = 150 Z-score = 5.23 Mol # 28 Raw score = 173 Z-score = 5.72 Mol # 29 Raw score = 151 Z-score = 5.25 Mol # 30 Raw score = 131 Z-score = 4.82 Mol # 31 Raw score = 171 Z-score = 5.68 Mol # 32 Raw score = 157 Z-score = 5.38 Mol # 33 Raw score = 136 Z-score = 4.93 Mol # 34 Raw score = 170 Z-score = 5.66 Mol # 35 Raw score = 185 Z-score = 5.98 Mol # 36 Raw score = 155 Z-score = 5.33 Mol # 37 Raw score = 162 Z-score = 5.48 Mol # 38 Raw score = 162 Z-score = 5.48 Mol # 39 Raw score = 92 Z-score = 3.98 Mol # 40 Raw score = 171 Z-score = 5.68 ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Nr of residues in profile : ( 43)Sequence identity for these residues only: % Seq id mol # 1 -> 100.0 30.2 32.6 37.2 32.6 32.6 32.6 32.6 32.6 32.6 27.9 23.3 23.3 20.9 23.3 18.6 23.3 23.3 23.3 18.6 16.3 4.7 11.6 16.3 7.0 16.3 11.6 7.0 11.6 14.0 9.3 11.6 11.6 9.3 55.8 11.6 11.6 11.6 11.6 30.2 % Seq id mol # 2 -> 30.2 100.0 58.1 48.8 58.1 51.2 53.5 58.1 55.8 37.2 25.6 9.3 16.3 16.3 16.3 16.3 16.3 16.3 16.3 16.3 11.6 16.3 14.0 14.0 16.3 18.6 16.3 14.0 11.6 16.3 11.6 9.3 16.3 14.0 34.9 11.6 11.6 14.0 11.6 27.9
[...]
% Seq id mol # 40 -> 30.2 27.9 41.9 39.5 41.9 44.2 44.2 41.9 41.9 39.5 23.3 20.9 16.3 23.3 18.6 14.0 16.3 14.0 14.0 16.3 14.0 9.3 11.6 9.3 9.3 16.3 11.6 9.3 11.6 14.0 9.3 14.0 9.3 7.0 34.9 9.3 9.3 9.3 18.6 100.0
Average sequence identity (%) : ( 27.961) St. dev. : ( 24.095) Minimum : ( 2.326) Maximum : ( 100.000)
Sum of maximum random scores : ( 746) Sum AVE+3SIGMA random scores : ( 243)
Score for molecule 1 = 513 Score for molecule 2 = 487 Score for molecule 3 = 617 Score for molecule 4 = 589 Score for molecule 5 = 617 Score for molecule 6 = 615 Score for molecule 7 = 626 Score for molecule 8 = 617 Score for molecule 9 = 592 Score for molecule 10 = 518 Score for molecule 11 = 454 Score for molecule 12 = 549 Score for molecule 13 = 516 Score for molecule 14 = 576 Score for molecule 15 = 586 Score for molecule 16 = 544 Score for molecule 17 = 561 Score for molecule 18 = 597 Score for molecule 19 = 597 Score for molecule 20 = 492 Score for molecule 21 = 596 Score for molecule 22 = 629 Score for molecule 23 = 564 Score for molecule 24 = 665 Score for molecule 25 = 602 Score for molecule 26 = 620 Score for molecule 27 = 600 Score for molecule 28 = 626 Score for molecule 29 = 617 Score for molecule 30 = 636 Score for molecule 31 = 640 Score for molecule 32 = 537 Score for molecule 33 = 632 Score for molecule 34 = 606 Score for molecule 35 = 553 Score for molecule 36 = 608 Score for molecule 37 = 610 Score for molecule 38 = 586 Score for molecule 39 = 435 Score for molecule 40 = 526
Minimum raw score : ( 400) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Now scan the database again with the new profile to see what comes up:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ../pftools/pfsearch -a hits.prf /nfs/scr_uu5/gerard/sprot34.dat | & tee pfs_hits.log ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Sorting the resulting file and running ZPROF again yields:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Working ...Nr of sequences scored : ( 59021) Average : ( 178.673) St.dev. : ( 44.654) Minimum : ( 26.000) Maximum : ( 594.000)
Remove "outliers" and re-calc ... Nr of sequences left : ( 58969) Average : ( 178.394) St.dev. : ( 43.636) Minimum : ( 26.000) Maximum : ( 355.000)
Remove "outliers" and re-calc ... Nr of sequences left : ( 58967) Average : ( 178.388) St.dev. : ( 43.625) Minimum : ( 26.000) Maximum : ( 351.000)
Remove "outliers" and re-calc ... Nr of sequences left : ( 58967) Average : ( 178.388) St.dev. : ( 43.625) Minimum : ( 26.000) Maximum : ( 351.000)
Converged !
MA /NORMALIZATION: MODE=1; FUNCTION=LINEAR; MA R1= -4.08913898; R2= 0.02292266; TEXT ='Z-score'; MA /CUT_OFF: LEVEL=0; SCORE= 353; N_SCORE= 4.00000000; MODE=1;
Z-score of 0 requires raw score 178 Z-score of 1 requires raw score 222 Z-score of 2 requires raw score 266 Z-score of 3 requires raw score 309 Z-score of 4 requires raw score 353 Z-score of 5 requires raw score 397 Z-score of 6 requires raw score 440 Z-score of 7 requires raw score 484 Z-score of 8 requires raw score 527 Z-score of 9 requires raw score 571 Z-score of 10 requires raw score 615 Z-score of 11 requires raw score 658 Z-score of 12 requires raw score 702 Z-score of 13 requires raw score 746 Z-score of 14 requires raw score 789 Z-score of 15 requires raw score 833 Z-score of 16 requires raw score 876 Z-score of 17 requires raw score 920 Z-score of 18 requires raw score 964 Z-score of 19 requires raw score 1007 Z-score of 20 requires raw score 1051 Z-score of 21 requires raw score 1095 Z-score of 22 requires raw score 1138 Z-score of 23 requires raw score 1182 Z-score of 24 requires raw score 1225 Z-score of 25 requires raw score 1269
1 9.53 594 P29257|LEC2_CYTSC 2-ACETAMIDO-2-DEOXY-D-GALACTOSE-BINDING SEED LECTIN II 2 9.21 580 P45797|GUB_BACPO BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 3 9.00 571 P27051|GUB_BACLI BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 4 9.00 571 P07980|GUB_BACAM BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 5 9.00 571 P04957|GUB_BACSU BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 6 8.95 569 P23904|GUB_BACMA BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 7 8.79 562 P02874|LEC_ONOVI LECTIN. 8 8.79 562 P02867|LEC_PEA LECTIN PRECURSOR. 9 8.77 561 P04122|LECB_LATOC LECTIN BETA-1 AND BETA-2 CHAINS. 10 8.66 556 P05046|LEC_SOYBN LECTIN PRECURSOR. 11 8.54 551 P16349|LEC_LATSP LECTIN. 12 8.47 548 Q01806|LEC1_MEDTR LECTIN 1 PRECURSOR. 13 8.45 547 P37073|GUB_BACBR BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 14 8.43 546 P29716|GUB_CLOTM BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 15 8.43 546 P16404|LEC_ERYCO LECTIN PRECURSOR. 16 8.40 545 P22973|LEC2_ULEEU ANTI-H(O) LECTIN II (UEA-II). 17 8.13 533 P05087|PHAL_PHAVU LEUCOAGGLUTINATING PHYTOHEMAGGLUTININ PRECURSOR (PHA-L) 18 8.04 529 P05088|PHAE_PHAVU ERYTHROAGGLUTINATING PHYTOHEMAGGLUTININ PRECURSOR (PHA- 19 7.99 527 Q01807|LEC2_MEDTR TRUNCATED LECTIN 2 PRECURSOR. 20 7.99 527 P19588|LEC5_DOLBI LECTIN DB58 PRECURSOR. 21 7.99 527 P05045|LEC1_DOLBI SEED LECTIN SUBUNITS I AND II PRECURSOR. 22 7.97 526 P23558|LEC1_LABAL LECTIN I (SEED LECTIN ANTI-H(O)) (LAA-I). 23 7.97 526 P02871|LEC_VICFA FAVIN (LECTIN). 24 7.88 522 P02870|LEC_LENCU LECTIN. 25 7.76 517 P24146|LEC4_GRISI LECTIN IV (GS4). 26 7.76 517 P16030|LEC_BAUPU LECTIN PRECURSOR. 27 7.69 514 P24806|MER5_ARATH MERI-5 PROTEIN. 28 7.58 509 P15231|PHAM_PHAVU LEUCOAGGLUTINATING PHYTOHEMAGGLUTININ PRECURSOR (PHA-L) 29 7.53 507 P22972|LEC1_ULEEU ANTI-H(O) LECTIN I (UEA-I). 30 7.28 496 P53301|YG46_YEAST HYPOTHETICAL 52.8 KD PROTEIN IN BUB1-HIP1 INTERGENIC RE 31 7.10 488 P35694|BRU1_SOYBN BRASSINOSTEROID-REGULATED PROTEIN BRU1. 32 6.96 482 P02873|LEC_PHAVU LECTIN PRECURSOR (ALPHA-AMYLASE INHIBITOR). 33 6.59 466 P17989|GUB_FIBSU BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 34 6.50 462 P07981|GUN1_TRIRE ENDOGLUCANASE EG-1 PRECURSOR (EC 3.2.1.4) (ENDO-1,4-BET 35 6.41 458 P33693|EXOK_RHIME SUCCINOGLYCAN BIOSYNTHESIS PROTEIN EXOK. 36 6.32 454 P38676|GUX1_NEUCR EXOGLUCANASE 1 PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 37 6.27 452 P19355|GUX1_TRIVI EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 38 6.27 452 P00725|GUX1_TRIRE EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 39 6.20 449 P38662|LECA_DOLLA LECTIN. 40 6.20 449 P19664|LEC_LOTTE ANTI-H(O) LECTIN (LTA). 41 6.04 442 P15828|GUX1_HUMGR EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 42 5.72 428 P45798|GUB_RHOMR BETA-GLUCANASE PRECURSOR (EC 3.2.1.73) (ENDO-BETA-1,3-1, 43 5.70 427 P13860|GUX1_PHACH EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 44 5.49 418 P42088|LEC_BOWMI LECTIN (AGGLUTININ) (BMA). 45 5.29 409 P19329|ARC1_PHAVU ARCELIN-1 PRECURSOR. 46 5.29 409 P02872|LECG_ARAHY GALACTOSE-BINDING LECTIN PRECURSOR (AGGLUTININ) (PNA). 47 5.13 402 Q06886|GUX1_PENJA EXOGLUCANASE I PRECURSOR (EC 3.2.1.91) (EXOCELLOBIOHYDR 48 5.08 400 P46237|GUNC_FUSOX PUTATIVE ENDOGLUCANASE TYPE C PRECURSOR (EC 3.2.1.4) (E 49 4.90 392 P19330|ARC2_PHAVU ARCELIN-2 PRECURSOR. 50 4.85 390 P32623|UTR2_YEAST UTR2 PROTEIN (UNKNOWN TRANSCRIPT 2 PROTEIN). 51 4.64 381 P39795|TREC_BACSU TREHALOSE-6-PHOSPHATE HYDROLASE (EC 3.2.1.93) (ALPHA,AL 52 4.48 374 P16270|LECN_PEA NONSEED LECTIN PRECURSOR. 53 4.05 355 P14894|CONA_CANGL CONCANAVALIN A PRECURSOR (CON A). 54 4.05 355 P02866|CONA_CANEN CONCANAVALIN A PRECURSOR (CON A). 55 3.96 351 P49254|CRP_CAVPO C-REACTIVE PROTEIN PRECURSOR.
Z-score cut-off : ( 4.000) Nr of "hits" : ( 54) % of database : ( 0.091) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Now it seems that virtually all sequences with Z > 4 are real hits (true positives).
The following papers may be of interest:
* Bairoch, A. (1993). The PROSITE dictionary of sites and patterns
in proteins, its current status. Nucl. Acids Res. 21, 3097-3103.
* Bairoch, A. and Bucher, P. (1994). PROSITE: recent developments.
Nucl. Acids Res. 22, 3583-3589.
* Gonnet, G.H., Cohen, M.A. and Benner, S.A. (1992). Exhaustive
matching of the entire protein sequence database. Science 256,
1443-1445.
* Gribskov, M., McLachlan, A.D. and Eisenberg, D. (1987). Profile
analysis: detection of distantly related proteins. Proc. Natl.
Acad. Sci. USA 84, 4355-4358.
* Gribskov, M., Lüthy, R. and Eisenberg, D. (1990). Profile analysis.
Meth. Enzymol. 183, 146-159.
* Henikoff, S. and Henikoff, J.G. (1992). Amino acid substitution
matrices from protein blocks. Proc. Natl. Acad. Sci. USA 89,
10915-10919.
* Lüthy, R., Xenarios, I. and Bucher, P. (1994). Improving the
sensitivity of the sequence profile method. Prot. Sci. 3, 139-146.
* Sibbald, P.R. and Argos, P. (1990). Weighting aligned protein or
nucleic acid sequences to correct for unequal representation.
J. Mol. Biol. 216, 813-818.
Created at Fri Dec 18 19:42:29 1998
by MAN2HTML version 971024/1.6