<!------------Discussion----------->
In conjunction with other tracks and specialized prediction tools, 
patterns of hydrophobic residues can sometimes be interpreted structurally. 
Smoothing the track over a rolling window size of 6 is effective in locating 
surface-exposed (polar) regions, whereas a window size of 20 is better suited for 
defining transmembrane domains that peak at a local average hydrophobicity 
score of 1.6 or more. Interior hydrophobic 
residues, while often packed tightly and critical to protein folding and 
stability, may be dispersed along the linear sequence. UniProtKB offers 
<A HREF="http://www.expasy.ch/cgi-bin/protscale.pl" TARGET=_blank>22 other choices</A> of 
hydrophobicity scales and considerable control over smoothing the
convolutions. 
<P>Using the genome-wide amino acid composition of human proteins, the
averaged 
amino acid has a hydrophobicity close to zero (-0.018). This reflects the 
design of the scale, the need for surface polar and charged residues to 
maintain solubility, and the frequencies of cytoplasmic and membrane-bound 
proteins. The track uses a baseline of zero. 
<P>
The mean hydrophobicity of the selected protein illustrated in the
Hydrophobicity histogram is not always informative, because transmembrane 
domains can be offset by highly polar regions elsewhere, yielding an overall 
mundane value. Unusual values may be apparent in the Hydrophobicity track, and 
will also surface as amino acid compositional anomalies. 
<P>
<P>
<H4>Kyte-Doolittle hydrophobicity scale</H4>
<BLOCKQUOTE>
<TABLE cellSpacing=2 cellPadding=1>
    <TR><TD vAlign=top>
<PRE>
 J. Mol. Biol. 157:105-132(1982)
 Ile:	+4.5
 Val:	+4.2
 Leu:	+3.8
 Phe:	+2.8
 Cys:	+2.5
 Met:	+1.9
 Ala:	+1.8
 Gly:	-0.4
 Thr:	-0.7
 Ser:	-0.8
 Trp:	-0.9
 Tyr:	-1.3
 Pro:	-1.6
 His:	-3.2
 Asn:	-3.5
 Asp:	-3.5
 Gln:	-3.5
 Glu:	-3.5
 Lys:	-3.9
 Arg:	-4.5
</PRE>
	</TD>
    </TR>
</TABLE>
</BLOCKQUOTE>
<CENTER>
<IMG height=262 src="hydro.jpg" width=286>
</CENTER>

