Hm. I suspect that your "more than 90% of the small string is included in the bi[g] string." can not be arrived at by 'wildcarding' some number of characters in your small string -- as is implied by your "with some letters of it missing..." suggestion.

It can only be achieved by breaking your small string into chunks and locating each (or most) of those chunks within the bigstring disregarding order.

Thus, I arrived at the following code which finds the entirety of the small string (100% of it) in the bigstring, in six discrete chunks out of order:

#! perl -slw
use strict;

my $bigstring="MNRIYSLRYSAVARGFIAVSEFARKCVHKSVRRLCFPVLLLIPVLFSAGSLAGTV
+NNELGYQLFRDFAENKGMFRPGATNIAIYNKQGEFVGTLDKAAMPDFSAVDSEIGVATLINPQYIASVK
+HNGGYTNVSFGDGENRYNIVDRNNAPSLDFHAPRLDKLVTEVAPTAVTAQGAVAGAYLDKERYPVFYRL
+GSGTQYIKDSNGQLTKMGGAYSWLTGGTVGSLSSYQNGEMISTSSGLVFDYKLNGAMPIYGEAGDSGSP
+LFAFDTVQNKWVLVGVLTAGNGAGGRGNNWAVIPLDFIGQKFNEDNDAPVTFRTSEGGALEWSFNSSTG
+AGALTQGTTTYAMHGQQGNDLNAGKNLIFQGQNGQINLKDSVSQGAGSLTFRDNYTVTTSNGSTWTGAG
+IVVDNGVSVNWQVNGVKGDNLHKIGEGTLTVQGTGINEGGLKVGDGKVVLNQQADNKGQVQAFSSVNIA
+SGRPTVVLTDERQVNPDTVSWGYRGGTLDVNGNSLTFHQLKAADYGAVLANNVDKRATITLDYALRADK
+VALNGWSESGKGTAGNLYKYNNPYTNTTDYFILKQSTYGYFPTDQSSNATWEFVGHSQGDAQKLVADRF
+NTAGYLFHGQLKGNLNVDNRLPEGVTGALVMDGAADISGTFTQENGRLTLQGHPVIHAYNTQSVADKLA
+ASGDHSVLTQPTSFSQEDWENRSFTFDRLSLKNTDFGLGRNATLNTTIQADNSSVTLGDSRVFIDKNDG
+QGTAFTLEEGTSVATKDADKSVFNGTVNLDNQSVLNINDIFNGGIQANNSTVNISSDSAVLGNSTLTST
+ALNLNKGANALASQSFVSDGPVNISDATLSLNSRPDEVSHTLLPVYDYAGSWNLKGDDARLNVGPYSML
+SGNINVQDKGTVTLGGEGELSPDLTLQNQMLYSLFNGYRNIWSGSLNAPDATVSMTDTQWSMNGNSTAG
+NMKLNRTIVGFNGGTSPFTTLTTDNLDAVQSAFVMRTDLNKADKLVINKSATGHDNSIWVNFLKKPSNK
+DTLDIPLVSAPEATADNLFRASTRVVGFSDVTPILSVRKEDGKKEWVLDGYQVARNDGQGKAAATFMHI
+SYNNFITEVNNLNKRMGDLRDINGEAGTWVRLLNGSGSADGGFTDHYTLLQMGADRKHELGSMDLFTGV
+MATYTDTDASADLYSGKTKSWGGGFYASGLFRSGAYFDVIAKYIHNENKYDLNFAGAGKQNFRSHSLYA
+GAEVGYRYHLTDTTFVEPQAELVWGRLQGQTFNWNDSGMDVSMRRNSVNPLVGRTGVVSGKTFSGKDWS
+LTARAGLHYEFDLTDSADVHLKDAAGEHQINGRKDSRMLYGVGLNARFGDNTRLGLEVERSAFGKYNTD
+DAINANIRYSF";
my $smallstring="GTMARNDGQGKAAATFMHISYNNFITEVDNLNKRMGDLRDINGEAGTWVRLLN
+GSGSADGGFTDHYTLLQMGADRKHELGSMDLFTGVMATYTDTDASADLYSGKTKSWGGGFYASGLFRSG
+AYFDVIAKYIHNENKYDLNFAGAGKQNFRSHSLYAGAEVGYRYHLTDTTFVEPQAELVWGRLQGQTFNW
+NDSGMDVSMRRNSVNPLVGRTGVVSGKTFSGKDWSLTARAGLHYEFDLTDSADVHLKDAAGEHQINGRK
+DSRMLYGVGLNARFGDNTRLGLEVERSAFGKYNTDDAINANIRYSFLE";
my $lenBig = length $bigstring;
my $lenSmall = length $smallstring;

my $o = 0;
WHILE: while( $o < $lenSmall ) {
    my $p;
    for my $l ( reverse 1 .. ( $lenSmall - $o ) ) {
        my $ss = substr( $smallstring, $o, $l );

        if( $p = 1+index( $bigstring, $ss ) ) {
            print "Found '$ss'($o:$l) at $p";
            $o += $l;
            next WHILE;
        }
    }
    ++$o;
}
[download]

The output from which is:

C:\test>1070240.pl
Found 'GT'(0:2) at 53
Found 'MA'(2:2) at 1160
Found 'RNDGQGKAAATFMHISYNNFITEV'(4:24) at 1076
Found 'DNL'(28:3) at 419
Found 'NKRMGDLRDINGEAGTWVRLLNGSGSADGGFTDHYTLLQMGADRKHELGSMDLFTGVMATYTD
+TDASADLYSGKTKSWGGGFYASGLFRSGAYFDVIAKYIHNENKYDLNFAGAGKQNFRSHSLYAGAEVGY
+RYHLTDTTFVEPQAELVWGRLQGQTFNWNDSGMDVSMRRNSVNPLVGRTGVVSGKTFSGKDWSLTARAG
+LHYEFDLTDSADVHLKDAAGEHQINGRKDSRMLYGVGLNARFGDNTRLGLEVERSAFGKYNTDDAINAN
+IRYSF'(31:275) at 1103
Found 'LE'(306:2) at 322
[download]

You'll probably want to exclude several of the smaller chunks because they are found at offsets that overlap where the two larger chunks are located. And that may be where you get your 90% from.

Removing/ignoring small chunks that overlap larger chunks is relatively easy; until you start to think about what to do if a smaller chunk only partially overlaps a larger. Then things start to get complicated.