We have derived Physical-chemical property based descriptors (PCP descriptors, E1-E5) were deduced from a comprehensive list of 237 physical- chemical properties (Venkatarajan and Braun, 2001). The five components will in general be differently conserverd at each position of the protein family. MASIA measures the conseration by calculating standard deviation of the E1-E5 magnitude at each position. Relative entropy R is calculated between the observed distribution of E in a particular column (five equally spaced bins) and the natural frequency of amino acids.
eq 1.
is the observed fraction of the component i in the bin
b and 
is the corresponding background frequency. High
relative entropy values indicate a significant difference between the observed
frequencies of distributions in a columb and the a priori background distribution.
Motifs are defined as contiguous positions in a multiple alignment where redsidues are significantly conserved according to on of PCP descriptors, that is those sites where the relative entropy R of at least one component E is above the R-cutoff value. A minimum length cutoff (L-cutoff) restricts motifs that are smaller length than this, and G-cutoff restricts the maximum number of insignificant positions between two significant positions in a motif. The default values are already present.
The PCPM macro will work only with protein sequences and the mutiple alignment currently must be in CLUSTALW format. Protein family which are 100% conserved are not suitable to be analysed using this appilication. Multiple alignment with identical sequences or false positive sequences will result in biased weighting and hence wrong results. A stand-alone program with advanced datamining procedure is available at PCPMer homepage. This site will provide additional information. For more details discussion please refer to Mathural etal, 2003.
Some suggestion for setting parameters:
R-cutoff: Use 2.0-3.0 for highly conserved family and 0.75-2.0 for moderately conserved.
G-cutoff: Use 0 or 1 for highely conserved family and 2-4 for moderately conserved family.
L-cutoff: Use 4-7 for highly conserved family and 2-4 for moderately conserved family.
The output is for APE protein family. The PCPM macro detects 12 conserved motifs using the parameter (R-cutoff 1.25, G-cutoff 2, L-cutoff 4). Motifs are highlighted using alphabets (a-z) in blue color below the representative (sequence 1) sequence from the multiple alignment (in this case human APE). Hovering mouse over the sequence will give the sequence number. The output also highlights conserved residues in the motifs in red color. At the top of the output the magnitude, standard deviation of every residue position is color coded for those who are interested in fine details (not shown here).
MOTIF LIST # Motif a: 61 62 63 65 66 67 68 69 70 71 72 73 # Motif b: 88 89 90 91 92 93 94 95 97 # Motif c: 124 125 126 127 128 129 130 133 135 138 # Motif d: 144 146 149 151 153 154 155 157 160 161 164 166 # Motif e: 170 172 173 174 # Motif f: 180 183 184 187 # Motif g: 203 204 205 206 207 208 209 210 211 212 213 214 # Motif h: 230 231 234 235 236 239 242 # Motif i: 246 248 250 252 253 # Motif j: 263 264 265 266 268 269 # Motif k: 273 276 278 279 280 281 282 283 285 287 290 # Motif l: 305 306 307 308 309 310 311 # Motif 0: 77 83 104 112 113 218 222 225 257 294
MOTIF DETAILSAPE_H._sapiens PKRGKKGAVAEDGDELRTEPEAKKSKTAAKKNDKEAAGEGPALYEDPPDH Motifs APE_H._sapiens KTSPSGKPATLKICSWNVDGLRAWIKKKGLDWVKEEAPDILCLQETKCSE Motifs aaaaaaaaaaaaa bbbbbbbbbb APE_H._sapiens NKLPAELQELPGLSHQYWSAPSDKEGYSGVGLLSRQCPLKVSYGIGDEEH Motifs ccccccccccccccc ddddddd APE_H._sapiens DQEGRVIVAEFDSFVLVTAYVPNAGRGLVRLEYRQRWDEAFRKFLKGLAS Motifs dddddddddddddddd eeeee ffffffff APE_H._sapiens RKPLVLCGDLNVAHEEIDLRNPKGNKKNAGFTPQERQGFGELLQAVPLAD Motifs gggggggggggg hhhhhhhhhhhhh iiiii APE_H._sapiens SFRHLYPNTPYAYTFWTYMMNARSKNVGWRLDYFLLSHSLLPALCDSKIR Motifs iii jjjjjjj kkkkkkkkkkkkkkkkkk APE_H._sapiens SKALGSDHCPITLYLAL Motifs lllllll
INPUT PARAM Relative Entropy-Cutoff (R):1.25 Relative Length-Cutoff (L):4 Relative Gap-Cutoff (G):2
Created and maintained by Venkat Mathura, 2003
