[Res.] KLDLHQTTQDLVALFAKVTVEQDDALLGNQISRFNRLFGVAEIADELKARDGDQRTALLSLFEYPNQVRLQAAKLTLAVAPVKAREQLEAIVSSKWFPQA : chn 1
[Pred] --AAAAA*AAAAAAA*00000*0AAA**RA*AAA00000*AAAAAAAAAB0*-**AAAAAA00G0R*0AAAAAA0000000AAAAAAAAAAA***-0**0 : chn 1
[Accu] --89988566677663789864537633695999568863567577765345-534777853666749688754554566345678887665423-9557 : chn 1
[Mult] --_____2_______2_____2____22__2________2___________2-23___________2_________________________223-_22_ : chn 1
[Comp] --FFFFFc_______cFFFFFcF___cc_Fc___FFFFFc________FF_c-cc______FFFFFcF______FFFFF_F___________ccc-_ccF : chn 1
[Actu] --0RR80QAAAAAAAAAAAAAAAAAAABRRQAAAAAAAAAAAAAAAAABG0RRQQAAAAAAQB0R0QAAAAAAAABQAB0QAAAAAAAAAAAABR00PAA : chn 1

[Res.] + GDAGCLDLLDDGTFKPK : chn 1
[Pred] + RGQA*AAAAQBR000-- : chn 1
[Accu] + 555555543356589-- : chn 1
[Mult] + ____2__________-- : chn 1
[Comp] + FFF_c____F_FFF_-- : chn 1
[Actu] + AAAAAAAAAABQRG0-- : chn 1

Statistics
  Success rate          = 0.4867    (= #'_' / (#aa - #chn * 4))          : ./PDBdata/TBM_FM/2I9C.code
  Extended success rate = 0.6372    (= (#'_' + #'c') / (#aa - #chn * 4)) : ./PDBdata/TBM_FM/2I9C.code
  Coverage              = 0.9823    (= (#aa - #'-') / (#aa - #chn * 4))  : ./PDBdata/TBM_FM/2I9C.code
  Mult candidate rate   = 0.1504    (= #'*' / (#aa - #chn * 4))          : ./PDBdata/TBM_FM/2I9C.code
  Ave. accuracy         = 0.6380                                         : ./PDBdata/TBM_FM/2I9C.code

// Remarks
// (0) Genaral 
//     '-' means no entry in the frag_code.tbl table.
// (1) Statistics 
//     2 amino acids at both ends of chains are excluded from computation.
//     '#aa' denotes the number of all amino acids.
//     '#chn' denotes the number of all chains.
// (2) [pred] entry 
//     '*' means multiple candidates for the position.
// (3) [accu] entry 
//     Accuracy = #{occurrence of the predicted code in the conversion table} / 
//                           #{occurrence of the amino acid fragment in the table}. 
//     (In the case of multiple candidates, the occurence of one of them is considered.)
//     And 'n' means prediction accuracy is greater than or equal to n / 10.
// (4) [mult] entry 
//     '_' means single candidate.
//     'n' denotes the number of candidates.
//     'U' means the number of candidates is greater than 9.
// (5) [comp] entry 
//     '_' means successful prediction.
//     'F' means prediction failure.
//     'c' means the actual code is contained in the condidates.