ࡱ> G hbjbjَ %w]4 #"4 accc5l@ $$&d  W | d 4 # L4 ar& aaJa4 D .a Prediction of initiation sites for protein folding with -helix preferences Prediction of initiation sites for protein folding Davor Juretia, Ana Jeron ia and Damir Zucib aPhysics Dept., Faculty of Natural Sciences, Univ. of Split, N. Tesle 12, 21000 Split, Croatia. bFaculty of Electrical Engineering, Univ. of Osijek, Istarska 3, 31000 Osijek, Croatia. Correspondence: Davor Jureti Physics, Faculty of Natural Sciences, University of Split N. Tesle 12, HR-21000 Split, Croatia ABSTRACT Background and purpose: The formation of -helices is thought to direct folding in helical and partly helical proteins. Particularly stable helices may be able to retain information about early folding events. Our goal was to test this hypothesis and to develop fast software package for prediction of helix nucleation and folding initiation sites in protein sequence. Methods: A statistical procedure used in this work consists in evaluating folding initiation parameter for each residue in tested sequence by using middle-helix preference functions and geometric average of position specific middle-helix preferences. The best known crystallographic structures of soluble proteins served for the extraction of preference functions. Position specific frequencies of folding initiation sites in observed helices were collected from proteins with known sequence locations of initiation sites. Results and conclusions: The highest frequency of folding initiation sites is in the middle-helix to C-terminus region of experimentally determined helices. Therefore, initiation sites for protein folding are likely to serve as helix-start signals. Overall sequence maximum of our folding initiation parameter is found at the sequence position which belongs to known folding initiation site and to observed -helix in 84% and 95% of tested sequences respectively. Sequence maximum of this parameter is inside transmembrane helix span for 68% of integral membrane protein sequences of known structure. We developed the Web sever for fast prediction of possible helix nucleation and folding initiation sites at the address: http://pref.etfos.hr/helix-start. Key words: prediction, folding initiation, helix nucleation, helix preference, preference functions, web server INTRODUCTION When folding begins locally in the sequence it is described as hierarchic folding process (8, 46). Then local elements of regular secondary structure form rapidly and to a good degree persist in the native secondary structure (35). In particular, fast creation of stable -helices has been observed in peptides (7, 23, 38, 51) and in proteins (35, 40). Helix formation is thought to be energetically favored for main-chain atoms of all residues except proline and glycine (2). Still, some residues, such as alanine, have higher helix-forming preference (12, 13, 18, 38, 42, 43) and higher helix propagation parameter s of the statistical mechanics model for  helix formation (10, 53) than other residues. Also, some position specific steric, electrostatic and hydrophobic interactions are found to favor -helix formation both in peptides and in proteins (2, 3, 4, 5, 37). The discovery of helix stop signals (24, 45) supported the idea that local sequence interactions determine helix nucleation and helix boundaries in native protein structures. Folding initiation sites have been found in many proteins (2, 6, 8, 41, 55). The prediction of nucleation sites in the sequence where fast helix formation is favored has been recognized as an important goal in folding simulations of helical proteins (3, 35, 50). The hypothesis that -helices, as local secondary structure, are "seeds for folding" (45) has been tested recently (16). It was essential to select reliable helix fragments as good candidates for folding initiation sites. This was achieved by using neural network learning algorithm designed to discriminate between -helices and non- structures, and by associating reliable patterns with minimums in the entropy of output vector. An alternative procedure is to use high helix propensities to select helix fragments of interest. High helix propensities are related to minimums in conformational entropy change, which is an important factor favoring -helix formation (17, 20). An obvious choice are middle-helix propensities (34, 49). Middle-helix propensities of Kumar and Bansal (34) are used in this work for the extraction of corresponding preference functions (28) from the database of crystallographic protein structures with the best resolution. Geometric average of amino acid attributes can also help to predict specific folding motifs (36). Its advantage is that it gives equal weight to high and low attributes in the sequence. In this work we used geometric average of seven position-specific amino acid preferences in the middle-helix positions derived by Richardson and Richardson (49). We report the performance in predicting folding initiation sites in proteins by using the combined index that takes into account both the geometric average of position-specific middle-helix preferences and sequence specific helical preference evaluated with middle-helix preference functions. In discussion we point out that longer helices, even including transmembrane helices from membrane proteins, are often associated with high maximum for folding initiation index. Locating folding initiation sites in the whole protein and helix-start signals in observed longer helices may help toward goal of selecting helices with a crucial role in early folding history. METHODS Protein data sets To extract preference functions (28) we used the data set of 100 soluble proteins determined by X-ray analysis (1.7 resolution or better) and NMR. There was no more than 30% pairwise sequence identity among these proteins (54). Corresponding Brookhaven Protein Data Bank (1)(PDB) codes are listed below: 1aac, 1ads, 1aky, 1amm, 1arb, 1aru, 1ben_a, 1ben_b, 1bkf, 1bpi, 1cem, 1cka, 1cnr, 1cnv, 1cpc_a, 1cpc_b, 1cse_e, 1cse_i, 1ctj, 1cus, 1dad, 1edm, 1fus, 1hfc, 1ifc, 1igd, 1iro, 1isu, 1jbc, 1kap, 1lam, 1lit, 1lkk, 1luc_a, 1luc_b, 1mct_a, 1mct_i, 1mla, 1mrj, 1nfp, 1nif, 1osa, 1phb, 1php, 1plc, 1poa, 1ppn, 1ppt, 1ptf, 1ptx, 1rcf, 1ra9, 1rge, 1rie, 1rro, 1sgp_e, 1sgp_i, 1smd, 1sri, 1snc, 1tca, 1utg, 1vcc, 1whi, 1xic, 1xso, 1xyz, 256b, 2ayh, 2cba, 2cpl, 2ctc, 2end, 2er7, 2erl, 2hft, 2ihl, 2ilk, 2mbw, 2mhr, 2mcm, 2olb, 2phy, 2rhe, 2rn2, 2sil, 2trx, 2wrp, 3b5c, 3chy, 3ebx, 3lzm, 3pte, 3sdh, 4fgf, 4ptp, 5p21, 8abp, 8ruc_a, 8ruc_i. The data set of 19 proteins with suggested sequence locations of folding initiation sites (16) is given in the Table 1 and 2. The data set of 31 sequences from integral membrane proteins of known crystallographic structure (30) is given in the Table 4. Correlation of folding initiation sites with position of residues in -helices The data set of 37 folding initiation sites with corresponding -helices (Table 1) has been prepared by us based on published data (9, 14, 22, 25, 26, 27, 41, 47, 50, 52) for above mentioned 19 proteins. The Compiani et al. (16) choice for folding initiation segments is very similar to our choice. Sequence location of secondary structure segments is taken from the most recent PDB assignment. Folding initiation frequencies are calculated as frequencies of residues suggested to initiate folding at specific helix positions. Table 1. Folding initiation sites and corresponding -helices. Protein data base considered (PDB codes are used) is the same one that Compiani at al. used (16). # protein initiation site corresponding (-helix position segment position 1) 1rhd 227-236 ELRAMFEAKK 224-235 2) 1phh 236-245 DERFWTELKA 237-245 3) 1cpv 9-20 ADIAAALEACKA 8-17 63-72 LKLFLQNFKA 60-70 4) 1gd1_o 104-113 DAAKHLEAGA 103-111 192-201 KDLRRARAAA - 257-266 NAALKAAAEG 252-265 5) 1gox 7-15 NEYEAIAKQ 8-16 142-149 RRAERAGF 134-146 330-338 MRDEFELTM 309-341 6) 2ts1 4-8 LAELQ 2-10 279-290 EALEQELREAPE 275-287 7) 3grs 37-44 RRAAELGA 29-43 8) 451c 40-47 AEAELAQR 40-49 9) 2ccy_a 39-50 DAAQRAENMAMV 40-58 91-98 TESTKLAA 79-102 10) 2cro 3-13 TLSERLKKRRI 3-14 11) 4mdh 160-167 NRAKAQIA 155-171 12) 8adh 333-342 ADFMAKKFAL 323-339 13) 7rsa 2-13 ETAAAKFERQHM 3-13 25-36 YCNQMMKSRNLT 24-34 14) 1hfx 26-31 WLCIIF 23-34 89-98 IMCVKKILDI 86-98 15) 2ci2 31-40 SVEEAKKVIL 31-43 16) 2mm1 9-17 LVLNVWGKV 4-17 29-33 LIRLF 21-35 102-115 KYLEFISECIIQVL 102-118 133-143 KALELFRKDMA 125-148 17) 1a2p 10-18 VADYLQTYH 7-17 25-36 ITKSEAQALGWV 27-33 45-49 VAPG - 18) 2lza 8-13 LAAAMK 5-14 28-36 WVCAAKFES 25-35 92-99 VNCAKKIV 89-100 19) 1hrc 7-15 KKIFVQKCA 3-10 64-70 LMEYLEN 61-69 91-101 REDLIAYLKKA 88-101 Table 2. Maximal folding initiation parameters for 19 soluble proteins (see Table 1 legend). Observed secondary structure (SS) is 'H' for -helix, 'B' for -sheet, 'U' for undefined, coil or turn structure and 'F' for folding initiation site (usually in the -helix conformation). The heptad segments are centered at amino acid number (AA) with sequence maximum for the combined index (FIP). Heptad maximums are found with geometric average of middle-helix preferences (49), while preference maximums are found by evaluating preference functions extracted with Kumar and Bansal (34) middle-helix preferences. Protein AA heptade SS AA preference SS AA combined SS heptad max max index segment max 1rhd 230 2.178 F 232 2.163 F 230 3.666 F ELRAMFE 1phh 111 1.510 H 334 2.294 H 334 3.452 H ICLRRIW 1cpv 65 1.826 F 15 2.650 F 14 3.744 F IAAALEA 1gd1_o 258 1.792 F 199 2.540 F 261 3.810 F AALKAAA 1gox 297 1.547 H 368 2.772 U 142 3.641 F QLVRRAE 2ts1 169 1.868 H 418 2.634 U 314 3.836 H ALRQAIR 3grs 449 1.401 H 2 2.584 U 39 3.595 F ARRAAEL 451c 46 1.772 F 43 2.559 F 45 3.667 F AELAQRI 2ccy_a 98 1.601 H 45 2.647 F 46 3.770 F RAENMAM 2cro 52 1.477 H 71 2.806 U 12 3.615 F KKRRIAL 4mdh 99 1.660 U 165 2.466 F 165 3.691 F AKAQIAL 8adh 38 1.671 U 11 2.086 U 338 3.345 F FMAKKFA 7rsa 28 1.657 F 6 2.294 F 6 3.643 F TAAAKFE 1hfx 28 1.622 F 123 2.926 U 119 3.180 U EQWYCFA 2ci2 48 1.546 B 38 1.841 F 37 3.116 F EAKKVIL 2mm1 47 1.600 U 133 2.267 F 133 3.428 F AMNKALE 1a2p 16 1.572 F 1 2.512 U 32 2.853 F EAQALGW 2lza 10 2.188 F 9 2.409 F 10 3.810 F ELAAAMK 1hrc 96 1.676 F 97 2.116 F 96 3.420 F DLIAYLK Middle-helix preferences Position-specific preferences found by Richardson and Richardson (49) for N3, N4, N5, middle, C5, C4 and C3 position of -helix are here defined as position specific middle-helix preferences. The sliding window with seven amino acid residues scanned the sequence and each heptad score is calculated as the seventh root of the product of seven position-specific preferences. The heptad score is assigned to the middle residue in the scanning window. The result is used directly as the heptad score profile of the sequence and indirectly in the combined parameter (folding initiation parameter, see below). Middle-helix preferences of Kumar and Bansal (34) are not position specific. We used a single set of 20 preferences that Kumar and Bansal extracted from the middle section of helices longer than eight residues. The data set of 100 soluble proteins served to extract corresponding middle-helix preference functions (28, 29). The evaluation of preference functions in tested sequence produced sequence dependent conformational preferences (28). In the combined index heptad scores are smoothed as three point averages and added to sequence dependent helical preference. In the following text the term combined index or folding initiation parameter (FIP) is used for the combination of Richardson & Richardson and Kumar & Bansal preferences as described above. The sequence location of transmembrane helices is predicted by using the SPLIT 3.5 suite of algorithms (29, 30) available at our Web server: http://pref.etfos.hr/split. Computer program for the calculation of folding initiation parameters was written in FORTRAN 77. It was translated into ANSI C and wrapped into the Web server HELIX-START, written in HTML and Unix shell script language. A graphic library, created for the HELIX-START server, enables fast (in seconds) graphical presentation of calculated profiles by using the server at the address: http://pref.etfos.hr/helix-start. Prediction accuracy calculations Prediction accuracy is reported as per-segment and per-residue accuracy. Only -helical segments longer than eight residues are considered for segment prediction of helices, but all residues observed in the helical conformation are considered for the per-residue prediction of -helix conformation. Prediction accuracy for folding initiation sites is calculated for the data set of 19 proteins in which such sites are known. To take into account overpredictions we report prediction accuracy as efficiency (# correct predictions/ # predictions), and as sensitivity (# correct predictions/ # observed features). Correct per-residue prediction is scored whenever higher then threshold FIP value is found inside observed feature. Overprediction is scored when higher then threshold value is found outside experimentally determined folding motifs that are being predicted. Per-segment prediction accuracy is calculated by taking into account only maximal FIP value inside observed segment. Higher than threshold FIP value, found during sequence scan outside observed segments, is scored as segment overprediction, if corresponding residue is a) at least three residues removed from the C-terminus end of observed segment, and b) at least eight residues removed from the residue already scored for segment overprediction. First and last eight residues in a protein are not considered for segment overprediction. RESULTS Prediction of folding initiation sites in soluble proteins Folding initiation sites in proteins are often found in the -helix structure (16, 50), which is not surprising because -helices are secondary structure elements known to fold very fast (35). Are folding initiation sites in helical proteins particularly strong helix-start signals, and if so are they found more often closer to helix N-terminus, helix-middle or helix C-terminus? The analysis of protein data set with known or suggested strong folding initiation sites revealed that frequencies of occurrence of folding initiation residues are maximal at middle-helix positions with a slight preference for the C-terminal half of the helix (Figure 1). Figure 1. Distribution of folding initiation sites in the helix span and for two residues external to helix span (N'', N', and C', C''). Folding initiation frequencies were calculated as frequencies of folding initiation sites at specific helix positions. In coiled coil helices (15) helix-start signals are likely to be hidden in each heptad of amino acid residues. If so, can we use a technique similar to Lupas et al. (36) to associate heptades having high score for middle-helix position with strong helix-start signals and potential folding initiation sites ? We used seven columns of position specific middle-helix preferences (49) to find such scores as described in the Methods section. We found that almost all folding initiation sites are indeed associated with at least one high heptad score, while highest heptad score for the whole sequence is often associated with the folding initiation site (columns 2, 3 and 4 in the Table 2). Since chosen Richardson & Richardson preferences were for middle helix positions we asked if preference functions derived from middle helix preferences (29, 30) are as good indicators of folding initiation sites as heptad scores. For 11 out of 19 proteins the highest helical preference in the whole sequence, evaluated with Richardson & Richardson preference functions (29,30), is found to be located inside folding initiation site. Out of 37 folding initiation segments in these proteins all but one are associated with a maximum in the sequence dependent -helix preference (not shown). Are these results reproducible with middle-helix preferences other than Richardson & Richardson s? To answer this question we used middle-helix preferences of Kumar and Bansal (34). Corresponding preference functions were equally good predictors of folding initiation sites (columns 5, 6 and 7 in the Table 2). Is combined index, defined as the sum of heptad score and sequence dependent -helix preference (evaluated with middle-helix preference functions) superior to above mentioned predictors? Sequence maximum in the combined index is associated with folding initiation site for 14 out of 19 sequences when middle-helix preferences are evaluated with Richardson & Richardson preference functions (29). Even better result for the combined index (Table 2) is achieved when middle-helix preferences are evaluated with Kumar and Bansal preference functions. Out of 19 sequence maximums for the combined index 16 are associated with folding initiation sites (Table 2) and 18 are associated with the observed -helix structure. Therefore, the term folding initiation parameter (FIP) seems to be appropriate for the combined index evaluated in this manner. What would be a good choice for the threshold FIP value suitable for predicting folding initiation sites? With FIP( 2.6 we find 230 out of 339 folding initiation residues and 31 out of 37 folding initiation segments listed in the Table 1. Segments are found by looking if maximal FIP value in the segment is greater or equal to 2.6. Lower FIP threshold increases the percentage of correct predictions (36 folding initiation segments are predicted with a threshold of 2.0), but produces too many overpredictions. It is possible of course that overpredictions of folding initiation sites are in fact correct predictions for helix-start signals that would uncover sequence location of some native helices. Prediction of the -helix conformation with the FIP index For the data set of 19 proteins where sequence location of -helices and folding initiation sites are both known it is possible to predict both features and to compare the prediction accuracy (Table 3). Correct prediction (prediction sensitivity) of 88% longer -helices is achieved with the FIP threshold greater or equal to 2.0. Even higher sensitivity of 97% for predicting folding initiation segments is accompanied with low prediction efficiency of 21%. Similar high percentage (82%) of correct prediction of longer helices is found for the data set of 100 soluble proteins (Methods) with the same FIP threshold. With increasing FIP threshold the number of correct predictions drops and becomes similar for folding initiation segments and for longer helices (Figure 2). This is not accidental. Helices that are not a part of the initiation sites are eliminated by setting a high threshold. For instance in the case of the highest threshold considered (3.6) a total of 12 correctly predicted folding initiation sites corresponds to 11 out of 12 correctly predicted helices, while only one helix is overpredicted (prediction efficiency of 92%). Table 3. Prediction accuracy for folding initiation sites and for -helices longer than eight residues. Protein data base:19 Proteins  100 proteinsPrediction accuracy* (%) Folding initiation sites (-helices(-helicesPer-segment Sensitivity 84 73 63 Efficiency 41 61 51Per-residue Sensitivity 68 41 31 Efficiency 24 73 63 * for the FIP threshold (see text) ( 2.6 Majority of helical residues are underpredicted. For instance in the case of 5288 helical residues out of 18769 in the data set of 100 soluble proteins the FIP value of 2.6 or greater correctly predicts 1635, overpredicts 953 and underpredicts 3653. Underprediction is less serious for initiation sites. Out of 339 such sites (residues) in 19 proteins 109 are underpredicted with the same FIP threshold. Figure 2. The dependence of the prediction accuracy on the chosen threshold for the folding initiation parameter (FIP). The ratio of correctly predicted to observed folding initiation segments (FIS, open circles) decreases, while the ratio of correctly predicted FIS to helical segments (plus symbols) increases with the increase in the FIP threshold. Table 4. Maximal heptad score and maximal folding initiation parameter (FIP) values for 31 sequences of membrane polypeptides with known sequence location of transmembrane helices (TMH). The letter "Y" in the last column denotes the case when the FIP sequence maximum is find inside the TMH span, while N denotes the case when this is not so. N (C) is the case when maximum is found at the protein C terminus. protein AA maximal SS AA maximal SS TMH heptade score combined index (FIP) 1prc_h 23 1.422 H 17 3.483 H Y 1aig_h 176 1.843 U 256 3.822 U N(C) 1prc_l 264 1.636 H 103 3.094 H Y 1aig_l 264 1.677 H 126 3.593 H Y 1prc_m 185 1.570 H 246 3.586 H N 1aig_m 216 1.541 H 248 3.758 H N 1kzu_a 27 1.284 H 31 3.121 H Y P04159 153 1.671 H 100 3.377 H Y 1ar1_a 99 1.760 H 339 3.741 H Y 1ar1_b 64 1.729 H 169 3.624 U N P06030 107 1.722 H 106 3.338 H Y 1occ_a 290 1.785 H 468 3.604 H Y 1occ_b 150 1.550 B 74 3.155 H Y 1occ_c 225 1.733 U 162 3.555 H Y 1occ_d 84 1.488 H 41 3.404 H N 1occ_g 16 1.332 H 17 2.406 H Y 1occ_i 62 1.462 H 15 3.404 H Y 1occ_j 10 1.516 H 9 3.385 H N 1occ_k 30 1.304 H 30 2.972 H Y 1occ_l 26 1.512 H 22 3.653 H Y 1occ_m 34 1.355 H 1 2.475 U N 1bcc_e 70 1.528 U 111 3.502 H N 1be3_k 15 1.420 U 14 2.764 U N 1bcc_j 53 1.602 H 32 3.145 H Y 1bcc_g 55 1.386 H 43 3.250 H Y 1bcc_d 63 1.633 H 63 3.684 H N 1be3_c 315 1.568 H 238 3.969 H Y 1brx 12 1.699 H 146 3.605 H Y 1afo 4 1.567 U 36 2.948 H Y 1bl8 90 1.474 H 7 3.382 H Y 1a91 13 1.442 H 13 3.786 H Y * PDB codes except for the Swiss-Prot codes P04159 and P06030. Folding initiation sites in membrane proteins For integral membrane proteins overall sequence maximum for folding initiation parameter is found inside observed transmembrane helix in 68% of sequences (Table 4). The percentage raises to 77% when remaining sequences are examined for sequence maximum of heptad scores. When extramembrane helices are taken into account as well then 87% of membrane protein sequences are found with maximal FIP inside some helix. For 31 membrane proteins (Table 4) 57% of residues are associated with the -helix conformation Examples of profiles for folding initiation parameter The profile of folding initiation parameter is shown for the cytochrome c (1hrc) (Figure 3). Three high maximums correspond to known folding initiation segments (26) (shaded columns for amino acids 7-15, 64-70 and 91-101) and to observed longer helices (bold line at the 0.5 level for amino acids Val 3 - Cys 14, Glu 61 - Glu 69 and Lys 88 - Asn 103). Hydrophobic-hydrophobic contacts with heme ligand are also found at these sequence positions (amino acids 10, 13, 14, 64, 67, 68, 94, 98). Another example of the FIP profile (Figure 4) is given for the apoprotein of the major light-harvesting complex of photosystem II in plant (Pisum sativum) (33). It has three FIP maximums that are almost as good indicators of the position of observed membrane spanning helices as Kyte-Doolittle preference functions (dotted line) (28, 29). It is also of interest that folding initiation maximums are associated with chlorophyll side chain ligands Glu 65, His 68 (helix B), Gln 131, Glu 139 (helix C) and Glu 180, Asn 183 (helix A) (notice that correct sequence numbers on the x-axis are obtained by addition of 25 N-terminal amino acids, omitted in the reported structure (33)).  Figure 3. The profile of folding initiation parameter for the cytochrome c (1hrc). Known folding initiation sites are shown as shaded columns up to the height of 1.0. Observed -helices are shown as the bold line at the 0.5 level.  Figure 4. The profile of folding initiation parameter (full thin line) and transmembrane helix preference (dotted line) for light-harvesting protein. Transmembrane helical preferences are evaluated from corresponding preference functions (29). The observed span of transmembrane helices A, B, C and of surface helix D is shown as the bold line at the 0.5 level. Predicted locations of transmembrane helices with the web server SPLIT (30) are shown as shaded rows up to the 1.0 level. Reported profiles are for the protein fragment 26-232. DISCUSSION Computational methods of sequence analysis with a goal to model folding process can profit from evidence that helix formation can direct folding. Presta and Rose (45) proposed that clusters of residues with high helix preference at the helix boundaries are necessary for helix formation during protein folding. Helix-start signals are expected to occur closer to the N-terminal helix positions (11), while helix-stop signals are often found at the helix N-terminus (24). However, helix-start signals, that can serve as folding initiation sites as well, are not located predominantly at the helix N-terminus (Figure 1). Folding initiation sites in 19 proteins that we considered are best associated with middle to C-terminus helical region. The frequency of folding initiation drops toward helix C-terminus and beyond, but it is still significantly higher than corresponding frequency for the N-terminal part of helix. Therefore, at least for one class of folding initiation sites in soluble proteins, helix nucleation of specific native helices initiates folding at or close to nascent middle-helix region Instead of taking into account all possible position-specific local interactions favoring helix formation we used position specific middle-helix preferences of Richardson and Richardson (49) and preference functions (28) calculated with middle-helix preferences of Kumar and Bansal (34). The justification for such a procedure for predicting folding initiation sites is a) calculated frequency of folding initiation sites which is maximal close to middle-helix positions (Figure 1) and b) expectation that helix formation dominates the folding kinetics of helical protein (35). Our procedure, which uses middle-helix preferences to calculate sequence profile of folding initiation index, may seem complicated, but the interpretation of the profile is straightforward. By choosing a high FIP threshold (3.6) one can identify those nascent helices that are crucial for initiation of protein folding. Prediction efficiency for finding such helices is higher than 90% when tested at limited data set of 19 soluble proteins with known folding initiation sites. About one third of observed folding initiation segments are then found with very few false-positive predictions. With a lower threshold (2.6) more than 80% of observed initiation segments are found and more than 70% of observed longer helices (Table 3). Still lower FIP threshold would produce even better prediction sensitivity, but decreased efficiency. For comparison the sensitivity of 34% was reported in predicting nucleation of protein helices with strip-of-helix hydrophobicity algorithm (48). Hem and chlorophyll ligands are known to promote helix formation (32, 44). Therefore, it is not surprising that amino acid contacts with such ligands in two sequences we examined are found to be associated with high folding initiation potential. Sequence span of observed transmembrane helices in integral membrane proteins is underpredicted when prediction is based on hydrophobicity analysis (29). While hydrophobicity, or helix preference based on hydrophobicity, is as a rule maximal in the middle region of membrane spanning helix this is not so for FIP maximums. FIP maximums are generally found closer to N or C terminus of transmembrane helices (Figure 4 and unpublished observations). Richardson & Richardson preference functions were used by us recently to refine the prediction of transmembrane helices and for the prediction of interface helices in membrane proteins (29, 30). The FIP index too has the potential to improve the prediction of transmembrane helix boundaries. Postulated folding mechanism of rapid hydrophobic collapse (39) has to be braked in those membrane proteins, whose entrance in the membrane requires partially unfolded structure (19, 21). The separation of early folding and hydrophobic domain in some of future membrane spanning domains is probably important for the membrane entry process. The light-harvesting protein enters thylakoid membrane from the stromal space so that hydrophobic C-terminals of helices A and B are oriented toward thylakoid space, while early folding N-terminal parts of these helices are oriented toward the stroma. If membrane entry is mediated by the translocase complex (31) than many charges in the N-terminal parts of helices A and B may assume specific configuration in the -helix conformation early in the folding history facilitating specific interactions with chlorophylls and with the translocation apparatus. In conclusion, for soluble helical or partly helical proteins, the initiation sites for protein folding correspond to sequence regions with strong middle-helix preference. By setting a high threshold for our FIP parameter one can select helices that initiate protein folding. In membrane proteins maximal FIP values are often associated with interface regions of transmembrane helices and can reveal topological signals (work in progress). ACKNOWLEDGEMENTS We thank Bono Lu i from Ruer Boakovi Institute in Zagreb who supplied us with copies of some papers needed for this work. Croatian Ministry of Science supported this research with the grant 177060. Thanks are also due to thoughtful comments by the anonymous reviewer. REFERENCES: 1. ABOLA E, BERNSTEIN FC, BRYANT SH, KOETZLE TF, WENG J 1987 Protein Data Bank. In "Crystallographic Databases-Information Content, Software Systems, Scientific Applications", Eds. ALLEN FH, BERGERHOFF G, SIEVERS R, pp 107-132, Data Commission of the International Union of Crystallography, Bonn, Cambridge, Chester. 2. AURORA R, CREAMER TP, SRINIVASAN R, ROSE GD 1997 Local interactions in protein folding: lessons from the -helix. J. Biol. Chem. 272: 1413-1416 3. AVBELJ F, MOULT J 1995a The conformation of folding initiation sites in proteins determined by computer simulation. Proteins: Struct. Funct. Genet. 23: 129-141 4. AVBELJ F, MOULT J 1995b Role of electrostatic screeining in determining protein main chain conformational preferences. Biochemistry 34: 755-764 5. AVBELJ F, FELE L 1998 Role of main-chain electrostatics, hydrophobic effect and side-chain conformational entropy in determining the secondary structure of proteins. J. Mol. Biol. 279: 665-684 6. BALDWIN, RL 1986 Seeding protein folding. Trends in Biochemical Sciences 11: 6-9 7. BALDWIN, RL 1995 -Helix formation by peptides of defined sequence. Biophys. Chem. 55: 127-135 8. BALDWIN RL, ROSE GD 1999 Is protein folding hierarchic? I. Local structure and peptide folding. Trends in Biochemical Sciences 24: 26-33 9. BYCROFT M, MATOUSCHEK A, KELLIS JT JR, SERRANO L, FERSHT AR 1990 Detection and characterization of a folding intermediate in barnase by NMR. Nature 346: 488-490 10. CHAKRABARTTY A, SCHELLMAN JA, BALDWIN RL 1991 Large differences in the helix propensities of alanine and glycine. Nature 351: 586-588 11. CHAKRABARTTY A, DOIG AJ, BALDWIN RL 1993 Helix capping propensities in peptides parallel those in proteins. Proc. Natl. Acad. Sci. U.S.A. 90: 11332-11336 12. CHOU PY, FASMAN GD 1974a Conformational parameters of amino acids in helical, -sheet, and random coil regions calculated from proteins. Biochemistry 13: 211-222 13. CHOU PY, FASMAN GD 1974b Prediction of protein conformation. Biochemistry 13: 222-245 14. CHYAN CL, WORMALD C, DOBSON CM, EVANS PA, BAUM J 1993 Structure and stability of the molten globule state of guinea-pig alpha-lactalbumin: a hydrogen exchange study. Biochemistry 32 : 5681-5691 15. COHEN C, PARRY DAD 1990 -Helical coiled coils and bundles: How to design an -helical protein. Proteins Struct. Funct. Genet. 7: 1-15 16. COMPIANI M, FARISELLI P, MARTELLI PL, CASADIO R 1998 An entropy criterion to detect minimally frustrated intermediates in native proteins. Proc. Natl. Acad. Sci. U.S.A. 95: 9290-9294 17. CREAMER TP, ROSE GD 1992 Sidechain entropy opposes helix formation but rationalizes experimentally determined helixforming propensities. Proc. Natl. Acad. Sci. U.S.A. 89: 59375941 18. CREAMER TP, ROSE GD 1994 -Helix forming propensities in peptides and proteins. Proteins Struct. Funct. Genet. 19: 85-97 19. DALBEY RE, ROBINSON C 1999 Protein translocation into and across the bacterial plasma membrane and the plant thylakoid membrane. Trends in Biochemical Sciences 24: 17-22 20. DOIG AJ, STERNBERG MJE 1995 Sidechain conformational entropy in protein folding. Protein Sci. 4: 2247-2251 21. EILERS M, SCHATZ G 1986 Binding of a specific ligand inhibits import of a purified precursor protein. Nature 322: 228-232 22. FERSHT AR 1995 Optimization of rates of protein folding: The nucleation-condensation mechanism and its implications. Proc. Natl. Acad. Sci. U.S.A. 92: 10869-10873 23. GRUENEWALD B, NICOLA CU, LUSTIG A, SCHWARZ G 1979 Kinetics of the helix-coil transition of a polypeptide with non-ionic side groups, derived from ultrasonic relaxation measurements. Biophys. Chem. 9: 137-147 24. HARPER ET, ROSE GD 1993 Helix stop signals in proteins and peptides: The capping box. Biochemistry 30: 7605-7609 25. HUGHSON FM, WRIGHT PE, BALDWIN RL 1990 Structural characterization of a partly folded apomyoglobin intermediate. Science 249 :1544-1548 26. JENG MF, ENGLANDER SW, ELOVE GA, WAND AJ, RODER H 1990 Structural description of acid-denatured cytochrome c by hydrogen exchange and 2D NMR. Biochemistry 29: 10433-10437 27. JENNINGS PA, WRIGHT PE 1993 Formation of a molten globule intermediate early in the kinetic folding pathway of apomyoglobin. Science 262: 892-896 28. JURETI D, ZUCI D, LU I B, TRINAJSTI N 1998a Preference functions for prediction of membrane-buried helices in integral membrane proteins. Computers Chem. 22: 279-294 29. JURETI D, LU IN A 1998b The preference functions method for predicting protein helical turns with membrane propensity. Journal of Chemical Information and Computer Sciences, 38: 575-585 30. JURETI D, JERON I A, ZUCI D 1999 Sequence analysis of membrane proteins with the web server SPLIT. Croatica Chemica Acta. In press. 31. KIM SJ, JANSSON S, HOFFMAN NE, ROBINSON C, MANT A 1999 Distinct "assisted" and "spontaneous" mechanisms for the insertion of polytopic chlorophyll-binding proteins into the thylakoid membrane. J. Biol. Chem. 274: 4715-4721 32. KRIGBAUM WR, KNUTTON SP 1973 Prediction of the amount of secondary structure in a globular protein from its amino acid composition. Proc. Natl. Acad. Sci. U.S.A. 70: 2809-2813 33. KHLBRANDT W, WANG DN, FUJIYOSHI Y 1994 Atomic model of plant light-harvesting complex by electron crystallography. Nature 367: 614-621 34. KUMAR S, BANSAL M 1998 Geometrical and sequence characteristics of -helices in globular proteins. Biophysical Journal 75: 1935-1944 35. LAURENTS DV, BALDWIN RL 1998 Protein folding: matching theory and experiment. Biophys. J. 75: 428-434 36. LUPAS A, VAN DYKE M, STOCK J 1991 Predicting coiled coils from protein sequences. Science 252: 1162-1164 37. LOCKHART DJ, KIM PS 1993 Electrostatic screening of charge and dipole interactions with the helix backbone. Science 260: 198202 38. MARQUSEE S, ROBBINS VH, BALDWIN RL 1989 Unusually stable helix formation in short alaninebased peptides. Proc. Natl. Acad. Sci. U.S.A. 86: 5286-5290 39. MATHESON RR, SCHERAGA HA 1978 A method for predicting nucleation sites for protein folding based on hydrophobic contacts. Macromolecules 11: 819-829 40. MATTHEWS CR 1993 Pathways of protein folding. Annu. Rev. Biochem. 62: 653-683 41. MOULT J, UNGER R 1991 An analysis of protein folding pathways. Biochemistry 30: 3816-3824 42. O'NEIL KT, DeGRADO WF 1990 A thermodynamic scale for the helix-forming tendencies of the commonly occurring amino acids. Science 250: 646-651 43. PACE CN, SCHOLTZ JM 1998 A helix propensity scale based on experimental studies of peptides and proteins. Biophys. J. 75: 422-427 44. PAULSEN H, FINKENZELLER B, KUHLEIN N 1993 Pigments induce folding of light-harvesting chlorophyll a/b-binding protein. Eur. J. Biochem. 215: 809-816 45. PRESTA LG, ROSE GD 1988 Helix signals in proteins. Science 240. 1632-1641 46. PTITSYN OB 1998 Protein folding: nucleation and compact intermediates. Biochemistry 63: 367-373 47. RADFORD SE, DOBSON CM, EVANS PA 1992 The folding of hen lysozyme involves partially structured intermediates and multiple pathways. Nature 358: 302-307 48. REYES VE, PHILLIPS L, HUMPHREYS RE, LEW RA 1989 Prediction of protein helices with a derivative of the strip-of-helix hydrophobicity algorithm. J. Biol. Chem. 264: 12854-12858 49. RICHARDSON JS, RICHARDSON DC 1988 Amino acid preferences for specific locations at the ends of  helices. Science 240: 1648-1652 50. ROOMAN MJ, KOCHER J-P A, WODAK SJ 1992 Extracting information on folding from the amino acid sequence: Accurate predictions for protein regions with preferred conformation in the absence of tertiary interactions. Biochemistry 31: 10226-10238 51. SCHOLTZ JM, BALDWIN RL 1992 The mechanism of -helix formation by peptides. Annu. Rev. Biophys. Biomol. Struct. 21: 95-118 52. UDGAONKAR JB, BALDWIN RL 1990 An early folding intermediate of ribonuclease A. Proc. Natl. Acad. Sci. U.S.A. 87: 8197-8201 53. WOJCIK J, ALTMANN KH, SCHERAGA HA 1990 Helix-coil stability constants for the naturally occurring amino acids in water. XXIV Half-cystine parameters from random poly(hydroxybutylglutamine-co-S-methylthio-L-cysteine. Biopolymers 30: 121-134 54. WORD JM, LOVELL SC, RICHARDSON JS, RICHARDSON DC 1999 Asparagine and glutamine: Using hydrogen atom contacts in the choice of side-chain amide orientation. J. Mol. Biol. 285: 1735-1747 55. WRIGHT PE, DYSEN J, LERNER RA 1988 Conformation of peptide fragments of proteins in aqueous solution: Implications for initiation of protein folding. Biochemistry 27: 7167-7175. PAGE  PAGE 22  EMBED Origin50.Graph   EMBED Origin50.Graph   EMBED Origin50.Graph   EMBED Origin50.Graph   (*FHhjlr04*q x } t}$$$$X%Y%))*_-~--/////n00022999R=S=CJCJ>*CJ>* ja5:OJQJ5:OJQJ 6OJQJOJQJ56OJQJ 5OJQJ6CJH*OJQJ6CJOJQJ56CJH*OJQJ56CJOJQJ CJ(OJQJOJQJ CJ0OJQJ8  lnp02 (dhdh$dhdh  lnp02 (q } qrst R#$$$$$$&(( )))**_-`-a-b-c-d-e-f-g-h-i-j-k-l-m-n-o-p-q-r-s-t-u-v-w-///n000:1z111|22293y333 cq } qrst R#$$$$$$&(( )$dh$dhdh )))**_-`-a-b-c-d-e-f-g-h-i-j-k-l-m-n-o-p-q-r-s-t-u-$dh$dhu-v-w-///n000:1z111|22293y33364s444 $d%d&d'd$d%d&d'd 9r $d%d&d'd$364s444555586x66677w777t88859u9999R=S=T=<>>>?a???3@y@@AKAAABcBBB5C{CCDD0EHxKMMMMM QhUiUjUkUlUmUnUoUpUqUrUsU{U|UUUYZ?]aci,{+,P4555586x66677w777t88859u9999R=S=T=<>>>?dh$dh $d%d&d'dS=T=<>>DDBDMMiU{UUUVYYZHjIj,m.mmOsXsttttuu u"uuuuuuvvv)vDvEv^vvvv x$x'xۺ· jmH OJQJmH ja5OJQJmH5OJQJmH;mH5;OJQJmHmH jOJQJ 6OJQJ 5OJQJjOJQJUhmHOJQJ56OJQJCJ5>*CJOJQJ5CJOJQJCJ1?a???3@y@@AKAAABcBBB5C{CCDDDDDFDHxKMM$dh$dh $d%d&d'dMMMM QhUiUjUkUlUmUnUoUpUqUrUsU{U|UUUYZ?]aci,m$dh$dhi,m.mmmOsPsttttuuu uvvvvvvvvvvvv(v)v5v6v7v:v;v*CJOJQJ5>*CJOJQJ5CJOJQJ 6OJQJOJQJ 5OJQJj5OJQJUhmH?yyyyyy*{+{{ ||||,}p}}}<~~~L\(l9ľ|vpjd^X 0  t      @        P        `      ,  p      ?            |,}p}}}<~~~L\(l9}IуY $d%d&d'd9}IуY!"PQ r268 ()*+,78}zwtqnkh    !      B C        S        c     (Y!"PQ r268 dh$dh$dh $d%d&d'd ()*+,78͘Θ}˧̧\]NP$$dh$dh8͘Θ}˧̧\]NPde+,Ĵƴ~~{xij34:; <ߵmn ab()ef;<HI Yjk#$ VW .Pde+,Ĵƴ~ոָ$ FHl~ضʸ$,M(ֿ -6unyz{(Rbf`s0J0^~ @R[s' CJOJQJ 5OJQJ 6OJQJ :OJQJOJQJYոָ /0Z[189df\^tv'( [\~[\ABϩЩ]:23'(ڰ۰[\۲ܲ0ָ /0Z[189df$\^tv'( [\01$01$%abXY+,HIefgh  lm%&ܜݜߞg$%uv017'k{b3V@O '() j#7U"jj> CJOJQJUVhmH"jvn> CJOJQJUVhmH jU <OJQJ0JmH0J j0JU CJOJQJ 6OJQJOJQJC$%abXYh&`#$dh $& #$$$+,HIedh $& #$$h&`#$)*,-DEFGIJabcdh jEU"j > CJOJQJUVhmH j?U"j > CJOJQJUVhmH jUefgh. 00P/ =!"#$%+0P/ =!"#$%+0P/ =!"#$%nn+0P/ =!"#$%+0P/ =!"#$%. 00P/ =!"#$%. 00P/ =!"#$%+0P/ =!"#$%+0P/ =!"#$%`!eRXW[#^cu6 [ 5axXklsهvM Am@(ZJZ`m NeDPDZPmH8"U[խQ#TU*?ΝxKf3={>k0ɘh3fuL!` (c\f:"G~4%5~PgǻS-y'ND0~F`p^ywOovAD0){S~Os'>Mx2Ӟұ|Mk]!2pBV(=TԏW8 k!EF8Pp]`z<㮍qWxÊwML[n,S=Ԗ|-I=nJ6{]gb䪇e%n(鸛 c%#)B;܄]f/FYgrY=cjl#E ӑ:HaW%x:Rsth<)1t'ӑ^-OG []6OG gӑNV(`i,OVbMPh¾R!S˄ŀ*BkkkrWqKH` ޘ lo6jwvo l ^F~c9UTՙz;>l8pPAi8pPAi8pPAi8pPAi8pPAi8pPA*M\5;V3b/W3ڢ5\""8jXB7fZ772[ُ`zj}zӬGF 5 sO؈8lkM1D(N?. ; V{2\`+gƞ6Y\cK3K8y(_h~|G1_ }]>*zEXsKПD/Vhcn%Ck6Z8FZDW#hC'n=j΍] K3D\Y֗>Qtkq-;GɸQ[9k*&Ϩy;)9MM4m;д@ӞSҴ4MsMW9?&Ҵp%M) M]4}Z4+ M$4|4&i;2C #M73饖Kρs4vR'Mg֕TzeP_nQJ<l=kw^Oo&Rm)Rπ7l<.tyRmsΓsv^68Toשqa |(8%["# ~I*:EI|U)MǡۙogM:=HYHX&9E|8{†ÒgjҨgU^i'lh&"<6wkwh]d*L~5\\w=?ȍE1<tg5_QUGh\ ]RކN)-=CNTî`U1=qmV]O:Q\PH̏64y^֞~ZvݝgaП^2:`5AK,;r${?K䩁l<ɪ%(wLC _La>`!p dXkH@Q2>x͙[hUϙf/7&^(%MְQX-Ii*!U_(VK B%(B0/ FB/A|(*%36$3w9s93 %&!b:?O?,!*JFK[~#"GՔ<*'1RUhź=] $#[R|.&z@]RݽMKN3EE۽vga~N[ucom%c w̴Z~pNNk[~Di ã]([LGYBO򾨗G@]AnpnW"+иVSBIK@I+8V.z lEsb2v@mA9}F l {xQv mۥ}@ y]Zpnվ(` 7@8Wr.eйHpnUA#QV xlLDI(ۘ xuLMMie;S)38qReҠ(arO CC1dX<EpZ_1 ZWVr7K"+|Y]!x簸 {bok:B|RTaznazv³.LϏ& OR\haw׫do'}j}(213s5ژ܋{owށPRLs9sf۽$bUhauj~tWv_AP}߲з.Tx} wHG'B}Q5:w5[n"L*(at3:rgt '#.EA@NF]5j 7SLda&r 9wBS6r2nFnFANF]s 3=d1crr2n!7̐[CNF]!w9+qd܍8r7 '#X9w"܉rIdH$rϦp)&Supif"'#ܥ|<×.km~( L sN9붑1h!haDܴܴG0ׯ#ׯcMCM\\z?8'1]ꗀ;ky]Z(i|ֆ|7I/"ʒm\b#YNQKQZV/VG5oɡqlhd644^G&Ĩz',/Q5/O=~WFGGOdR9CN{ |ÁE ~DzNQĈ5Y*w`!A׌ێlAuK+% =~ 0T8x՚hUs=#`i2KL\q&[2!+&CbeG !3鐹ve(Z]AVVkRZglúbІ+ocMɟz=~\o(q I?%Ĕ#|RGIR Р!/nݠU~mY_-pohΧvdE`ܾo=]{amfwfσό? 2)]8HY>d5fS,X1b[W6j4v9n*.gRfUA=(qA=s4- Ow\|d\.O5PCŏ̬[ [BtCoa+*|ѷ }+ <R|?jD&AYN{k m c\X3Ǹ1֌Qb\X3ڤĸIU֌֌faㆭb\X3ƅ5݈a܍%g51|0ŸC➟Gc91h=V#ŭq'j'ƥR7 QhFqQqVkwvc\z9%q ~qi 7uF[?t 0;^;)f_byby/KǬF1<ٗ}kVR>bQ>be)fG10ϣ}mt*\,>Ӧ}9^ =|C-nN(N(KY-0n@6 1jFjڭ.06G6Ǹ^rKX-aj VK36H߳v7F[|r/;aVw~l~W>)cKwo]{Y%FyEn9G.T]<*ȓrHG~<"'N@U>jӢv5&S⸪g:?T |B}\[*oR9oJEs Ϲw8gxsr~_WF;(wȸul{nvYgb}Xs7Kz=߳ ,NS(F/ w*(b0ߵkS, xc OቆO4w:^@-@@' ?n UQhJ ٖ#%'J|*Oh=G'=_F^TQCԎb;z}OH+O{<}Nfz_*^C't_#ߑ ~[W_𨤤df߄J4Z!gng/nnJJ"=y6o]}G5 MYV6/~sC+4MK8Οg;^<Z٧N~ice&e˄l۴l/{md-EmZv:y(}T~I^Smnu݃Ew)Ew}*[!m]|W%d"仒+q밭ЅV\g]u9/>m>"u.ù .ƹW+˕.ѹDi늅ue:)l]s[Wm]sօ E̺E9l>nٺtUurG1˳uY[ws)G/pvyv{v HHKHHّّّّ6г#-޳#-ͳ#-޳#mgGZgGZWzU}:߯}Ejs؎-2[Ÿ\i]u%J[W,+.ӹLabuuܺnRK.\-b-b.ι8fu֥;NsU]];Y;=[qt{9ZN ̣zlk`y U9[)V!*i&b.XVgcd]6U60|˓XlːNb=Peb"b[,^`^l#61K_ZX&|&y5\v Úmk5X>؝< 3~6|6$1/U|X<;x/GTq'g"CH *DJA$xo|SD- TQ]s"G,'Q,df%?1F)i|@eHr)ID Hy dU>r) iَy.iCOlSL`d%ZmQ|d6?( q%3K겢UBUUɲ ;Llo5Dɱ|l<9dʑzL͔tA%ȏTܩΊPS3EQ΃x`~Ip bavm0{ f6,X dlZ=gsㅦ!*(|K0oFt~4ɿ ?P>/.&Z,qeX4 -CGEP bP(N BKl>%ň9%CŊP1=2ʼnq&_E(^T ^H6 wQ-Y}HqsreyͣH.ٍB##HE:Ri\C:#ې$d2:. M;.] MC P&4MR{U@s@9:iZ­7+Kߠu: Q fI.=W)GК.wT۵^ڤ꡵@:zTG7CQZT hHu<fm:\YXzj!:j\KnU@@vޱ}(Zc*P|OqG߁6RuY5E(GEu524SW4MQ1p+B#Ap}U,εQ-@AZ8WWtHS-.ZB MܹSzз m[ʶ Md;2͝u @AwnHD.Uv i[7z4;R W[&.i.oH4F :tԹq;{G i^@@87Y4ܣG\4.z@A$X jعb0HSUH @~CЗ 3VCA4mvnZҴ䩠@f9z)ݹdJiM|$+HSGQ& MG.i+f~i wʹ<6dNv?2f}{b:= zIƳIAưpl T#N [ip ٣ MW9W= M;K~i;@ M{˥'AAwn =zḯIOizܜc74f gA49p9@ژ3:bh.{ H]AAf=/iz^s EAfrn zd0rU&;@YZ :2.2{Tdl%F 3;w#{d%;A:x&FenDYzdo-7.@o̵ѹ<!\W;VksmP \ϹuJХ p)_d7u5φKk@AfC] MS4 6Ȭe|isW M'#4F"tHSsb:!6niJp.El iJypMu:_|zisAAr+ MVo6 "@;W[4E;Rnij'w;A}@;* A#@qn(HSs .A^un zdD+{@4핟rtZETsu~P=pm!P[0\_Hun: :VquHLunW@({^@sAfrpUU MU5e M\wu3H;z뚫f M4UqSAsG#[Iأt̉ѺRqz͒}Aπ4e~pKЛ M+@i>@ULijS {@ Խp'e:H19K9tL> (< "6}LwGqPM:5WXVj;sfz4QݪګvjZ:j^FkMoRk@I M)jc0Wq )Umq/_1j\/ڋ~s0(i: ޮ:^,ZoT_{tR UR nWAʪH̺];++~u3Y dʠ nUȪAf./)qY@97@./Ck_Y{Z@f\ seПA2 =H)@ E="is+A4-WeAO4Ms 2gs5~ spDP[Ez M+/ʁ9IkAAqnoZ&. Ҵ;u!o;H-֠GyNU44Gk 2%@}xUy'P+XxPM 23:wŃŇκ`74;zV5ʹ؋IZk=yj|))9siS;ޤJk4HH sڎbi)^}Jϕ/m E70WI;Kэ45kOS_JэtJ'Hэ\2]RHi&#eFJ˛CAjnR+EW}>\~$UBBΕo8ΡiCC{rÃ<. k (뼠u!AF<)a> ԧueX--k6-6{(mnlc){DأүJ;5tCiI=4-'[J[{JZ&EP3ҫRt}{NT~iM)Ҷ?/m)EPڞi֟H=4>$%E jg_Β{(K|)Ƈ_n{(+3ėUZ!9k׵~#63,əٴm7m%gv-'9s ]Rovn8"1*͎/]-En]aI[r7,=Mz:@kK;(ck)H[~*9)njWmo?vkE '>əagH%g,#əuTo ,93n(nu8FzImԖuRə߱\rfE?O$ge4Crfms{̶`=!Ҟ3J nwl;Hwg{T-c6'JJliOR͞.͵gh˞HErf'oj[k%:iəjGnrO-zM:ީ,X:[rv|t5*ekZjv}tLꉊ{t=eq][.tnΕH,Kڽ>mi-#~{{=ET=[z@XΑf.*?k穵ٔtToWkJi=N%g~j/Q$gUK%gֵMKI|I KIμI+WHυWj}RU}3wҵ3WUKH90-Frax}+uxa $g֦}H_I|Am봧7JOI9_ϽmY*əh$+9sc%gEWrfڶ[m$g6͊/fJ.ۤ%g>-ސVcKI"əicwJK\L;+vt)ՎBGJGm/g^Iѧ}3֌H3$g^OO= '-y,KK%g jJHH<$]"9s!aTə>_ÂZ!E7`4WrffRiSəh6X&3hCə1ڰyTzI>cÒ39ZNOU{ЬYJ{|.(97'jK).93/&3[5FMrf[_IJI;9N &)U~򿤋%g.Lӓ7K+}[ic@zEr GALXrsX\ZCrf&mN,6ن{J{Hv8:Brj3$g΍P]&9sUlM6əbk=[[zTr)k3oVfII|E[+zJ[Kѭw|=i)uĝ)ugOBOGvE|}Jəkiwī;@{"d|CI?+HzSrjH_J|G['DZWrf}ͤFI8Gi֟K?WB:Nrj ,9s~_mk$gU'tݭhτKJŷI||jnv* 3NUk|]Uލ2X]Μ@KVvZufh=@+~!e-&$k`}B֯5Z`գFZjfmk ufi9g Aa*G>1rb|KVvZufh=@+~!e-&$k`}B֯5Z`գFZjfmk ufi9g Aa*GߒFZe-Њu6FY˴ 4Z~բk56ҚP3k X;h]5Kˡdx;1VWD4dzP/kVDC4ZMIV_kh(ZGG &BkC)Zfr(G@ô1Tb4u'ƿuxs-ZYiiɚ^Yh#hL@ U[jQVXiMֆR.͚P6Yic:iNWD4dzP/kVDC4ZMIV_kh(ZGG &BkC)Zfr(G@ô1Tb4u'ƿgʫrb|KVvZufh=@+~!e-&$k`}B֯5Z`գFZjfmk ufi9g Aa*GU91%Q+k;-:Y3ZQ?mih5UjZ-JQk# 5PօYʳ 0m X#M݉o`ߒFZe-Њu6FY˴ 4OB|jh(ZGG &BkC)Zfr(G@ô1Tb4u'ƿ;ϫrb|KVvZufh=@+~!e-&$k`}B֯5Z`գFZjfmk ufi9g Aa*G\U91%Q+k;-:Y3ZQ?mih5UjZ-JQk# 5PօYʳ 0m X#M݉oߒFZe-Њu6FY˴ 4XߪPk %XhքY[hm(ABݬYZYhhu6J㑦7zUNoINKN ; Y5?;1lJ٠%N[!,ft'5;䪐Ytfp ٴۢDPPY܋ЋE-x7 -tZXt?KiMKߙ#7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tń.&*M:j:h-کG.j" .RQi.jˣZJV:j-z:hn=:pv@WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg G'n麘ન66}nFꔛh3($JE ʪ9. j+Zꨡ.n&[蠞bb]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL[L8;h.WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg Gf3ܾUQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tń.&*M:j:h-کG.j" .RQi.jˣZJV:j-z:hn=:p@WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg Gt]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1聮 jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-&>u1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tu1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tż jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL[̿z~KWE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg GS/u1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tńoC t]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1-]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL[L@WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg G'6u1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tt]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1;]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL[L^@WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg GQ(.WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg G[.&*M:j:h-کG.j" .RQi.jˣZJV:j-z:hn=:?u1UQMmrmQ)7G umN=rgtQQpI i.jˣZJV:j-z:hn=:wo麘ન66}nFꔛh3($JE ʪ9. j+Zꨡ.n&[蠞bb*t]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1εi1jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-{t]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1񖮋 jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1[z:צM:j:h-کG.j" .RQi.jˣZJV:j-z:hn=:?zbڨ!Snۢz&(2(<*Ej⺨13nVۣz~ot]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1񖮋 jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)->u1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=t<{ jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-^bڨ!Snۢz&(2(<*Ej⺨13nVۣz~Wo麘ન66}nFꔛh3($JE ʪ9. j+Zꨡ.n&[蠞bb>-]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL?n1O<_+_^u=W◾7>xxxvt}g|7 zsu!_/qx?1ŋ7~퍯֯>+O?ׯ?|uS<~?8g^ۣ߽g?^?⇇o%^yW~#ϻz}ӻ_'}=KO??;(?tߏ~iw;?z>?;?'{Kg_?>q*ߧ⯽7KڇƷ>??GO_rL{}o|矕~~o/>7>k·-7A}x-LǗ'~./տs2&KO6|?Ou1_׾Zʛo/|ͷo~+卯!\~R[r1z%Sdÿj}?}{o^?Խxo1}vү<~OqE%4^裏>O>?}?X7>-%|Fxn{O/_L5'/Q5O#7Ddd" 0  # A26'Ŭ~r%`2{6n`!s6'Ŭ~r%`2fA~ OU(9A6xܚ Msu-$$&䒷%d˄ M` JLRyKQ4Rq!24 LHޓ1ܢfskw:^@-@@' ?n UQhJ ٖ#%'J|*Oh=G'=_F^TQCԎb;z}OH+O{<}Nfz_*^C't_#ߑ ~[W_𨤤df߄J4Z!gng/nnJJ"=y6o]}G5 MYV6/~sC+4MK8Οg;^<Z٧N~ice&e˄l۴l/{md-EmZv:y(}T~I^Smnu݃Ew)Ew}*[!m]|W%d"仒+q밭ЅV\g]u9/>m>"u.ù .ƹW+˕.ѹDi늅ue:)l]s[Wm]sօ E̺E9l>nٺtUurG1˳uY[ws)G/pvyv{v HHKHHّّّّ6г#-޳#-ͳ#-޳#mgGZgGZWzU}:߯}Ejs؎-2[Ÿ\i]u%J[W,+.ӹLabuuܺnRK.\-b-b.ι8fu֥;NsU]];Y;=[qt{9ZN ̣zlk`y U9[)V!*i&b.XVgcd]6U60|˓XlːNb=Peb"b[,^`^l#61K_ZX&|&y5\v Úmk5X>؝< 3~6|6$1/U|X<;x/GTq'g"CH *DJA$xo|SD- TQ]s"G,'Q,df%?1F)i|@eHr)ID Hy dU>r) iَy.iCOlSL`d%ZmQ|d6?( q%3K겢UBUUɲ ;Llo5Dɱ|l<9dʑzL͔tA%ȏTܩΊPS3EQ΃x`~Ip bavm0{ f6,X dlZ=gsㅦ!*(|K0oFt~4ɿ ?P>/.&Z,qeX4 -CGEP bP(N BKl>%ň9%CŊP1=2ʼnq&_E(^T ^H6 wQ-Y}HqsreyͣH.ٍB##HE:Ri\C:#ې$d2:. M;.] MC P&4MR{U@s@9:iZ­7+Kߠu: Q fI.=W)GК.wT۵^ڤ꡵@:zTG7CQZT hHu<fm:\YXzj!:j\KnU@@vޱ}(Zc*P|OqG߁6RuY5E(GEu524SW4MQ1p+B#Ap}U,εQ-@AZ8WWtHS-.ZB MܹSzз m[ʶ Md;2͝u @AwnHD.Uv i[7z4;R W[&.i.oH4F :tԹq;{G i^@@87Y4ܣG\4.z@A$X jعb0HSUH @~CЗ 3VCA4mvnZҴ䩠@f9z)ݹdJiM|$+HSGQ& MG.i+f~i wʹ<6dNv?2f}{b:= zIƳIAưpl T#N [ip ٣ MW9W= M;K~i;@ M{˥'AAwn =zḯIOizܜc74f gA49p9@ژ3:bh.{ H]AAf=/iz^s EAfrn zd0rU&;@YZ :2.2{Tdl%F 3;w#{d%;A:x&FenDYzdo-7.@o̵ѹ<!\W;VksmP \ϹuJХ p)_d7u5φKk@AfC] MS4 6Ȭe|isW M'#4F"tHSsb:!6niJp.El iJypMu:_|zisAAr+ MVo6 "@;W[4E;Rnij'w;A}@;* A#@qn(HSs .A^un zdD+{@4핟rtZETsu~P=pm!P[0\_Hun: :VquHLunW@({^@sAfrpUU MU5e M\wu3H;z뚫f M4UqSAsG#[Iأt̉ѺRqz͒}Aπ4e~pKЛ M+@i>@ULijS {@ Խp'e:H19K9tL> (< "6}LwGqPM:5WXVj;sfz4QݪګvjZ:j^FkMoRk@I M)jc0Wq )Umq/_1j\/ڋ~s0(i: ޮ:^,ZoT_{tR   !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~<Root Entry F\\uData  NWordDocumentObjectPool(\u_1041329782ac(@=((Ole ObjInfoContents> CPYA 4,210 #  $@X r ??0PBFIG3_AA. x ?@@@@@@ @"@$@&@(@*@,@.@0@1@2@3@4@5@6@7@8@9@:@;@<@=@>@?@@@@@A@A@B@B@C@C@D@D@E@E@F@F@G@G@H@H@I@I@J@J@K@K@L@L@M@M@N@N@O@O@P@@P@P@P@Q@@Q@Q@Q@R@@R@R@R@S@@S@S@S@T@@T@T@T@U@@U@U@U@V@@V@V@V@W@@W@W@W@X@@X@X@X@Y@@Y@Y@Y@Z@@Z@Z@Z@[@@[@[@[@\@@\@\@\@]@@]@]@]@^@@^@^@^@_@@_@_@_@`@ `@@`@``@`@`@`@`@a@ a@@a@`a@a@a@a@a@b@ b@@b@`b@b@b@b@b@c@ c@@c@`c@c@c@c@c@d@ d@@d@`d@d@d@d@d@e@ e@@e@`e@e@e@e@e@f@ f@@f@`f@f@f@f@f@g@ g@@g@`g@g@g@g@g@h@ h@@h@`h@h@h@h@h@i@ i@@i@`i@i@i@i@i@ r !`??PBFIG3_B r !`??PBFIG3_C r !`??PBFIG3_D r `??PBFIG3_TMH x Q?\(\?Q?Gz?(\?Q?Q?Q?Q?)\(?)\(?)\(?)\(?{Gz?{Gz?{Gz?{Gz?? ףp= ? ףp= ?)\(?Q?)\(?Q?p= ף?Q?Q?p= ף?Q?)\(??Q?{Gz?{Gz?{Gz?{Gz???Q?Q?(\?ףp= ?q= ףp?{Gz?RQ?)\(?Q?ffffff@(\@ףp= @Gz@333333 @ @Q @q= ףp@HzG@{Gz@HzG?Q?Q? ףp= ?Q?p= ף?(\?Q?HzG?Gz?(\?zG?ףp= ?ףp= ?p= ף?{Gz?(\?(\?(\?(\?(\?? ףp= ??Q?Q?ףp= ?{Gz?= ףp=?= ףp=?{Gz?ffffff?RQ?(\?p= ף?Q?Gz?ffffff?HzG?{Gz@= ףp=@(\@)\(@{Gz@Q @@ @333333 @@{Gz @)\( @@HzG@@Gz@zG?p= ף?RQ?Gz?Q?\(\?HzG?Q? ףp= ?Gz?Gz?(\??ףp= ??Q??Gz?)\(??RQ?RQ?Q?Q?ffffff?{Gz?Q?)\(?)\(? ףp= ?(\?{Gz? ףp= ? ףp= ?(\?(\? ףp= ?{Gz?{Gz?p= ף?Q?Q??{Gz? ףp= ?(\??)\(?Q?ףp= ?(\?333333@{Gz@p= ף @@\(\@\(\@333333 @Gz @(\@Gz@Q@Q?Gz?q= ףp?Q?= ףp=?Q?q= ףp?)\(?(\?(\? ףp= ??RQ?333333?p= ף?Q?? ףp= ?{Gz? ףp= ? ףp= ?Q?{Gz??zG??q= ףp?(\?Q?Q?q= ףp?HzG?? r `??PBFIG3_BET r `??PBFIG3_TUR r `??PBFIG3_UND r `??PBFIG3_H.T x {Gz?Q??Gz?Q?Q?HzG?(\??333333?= ףp=?ffffff?Q?= ףp=?Gz?Q?= ףp=?{Gz?Q?p= ף?Gz??Q?)\(?\(\?Q?ffffff?Q?(\?(\? ףp= ?\(\?Gz?(\?)\(@Q@HzG@p= ף@Q@\(\@(\@\(\@p= ף@HzG@(\@@Q@\(\@p= ף@(\@{Gz@Gz?Gz?Q?HzG?q= ףp?(\?HzG??333333??Gz?(\?{Gz??HzG??333333?\(\?(\@= ףp=@q= ףp@p= ף@@p= ף@ףp= ?Gz?Q?(\?(\?p= ף??333333?333333?(\?zG?ffffff? ףp= ?p= ף?333333?Q?ףp= ?)\(?Gz?\(\@@= ףp=@ ףp= @ @(\ @(\ @(\ @ffffff @RQ @Gz @RQ @ ףp= @@HzG@zG@Gz@Q@p= ף@{Gz@ףp= ??)\(?Q?zG?HzG??ffffff?\(\?Gz?(\?Q?(\?Gz?q= ףp?Gz??Gz?Gz?Gz?p= ף?RQ?)\(??{Gz?= ףp=?)\(??RQ?p= ף?333333?Gz@Q@(\@Q@Q @HzG@HzG @333333 @Q @= ףp=@ffffff@ ףp= @Gz@\(\@{Gz@333333@Gz@= ףp=@@Gz@333333@q= ףp@Q@(\@Q@(\@(\@Gz@Q?Q? ףp= ??)\(?(\??Q?(\?Q?zG?Q@333333@Gz?{Gz?Q?= ףp=?\(\?{Gz?(\?(\?\(\? ףp= ??ffffff?(\?Gz?ffffff??RQ?{Gz?zG?{Gz? r `??PBFIG3_J r `??PBFIG3_MOMA r `??PBFIG3_MOMB r `??PBFIG3_INDA r `??PBFIG3_INDB r `??PBFIG3_MBET r ` `??PBFIG3_DIG x ??????????????????????????????????????????????????????????????????????? r `??PBFIG3_OBS x ???????????????????????????????????????????????????????????????????????????????????????????????? r `??PBFIG3_SC1 r `??PBFIG3_T.60 s Plot1$ `  ` C:\ORIGIN Fr,,s 9 l@I@@-"@@? @yl "N 9 S2? ף= ף=Pd1 c FY  +b`ȿ&Y%ɿ>> L4  4 c FY  b`ȿ&Y%ɿ>> L3  3 c FY , qb`ȿ&Y%ɿ>> L2  2 c FY  b`ȿ&Y%ɿ>> L1  1 c / J 8?yE? ,>>  l "]  , TMH PREFERENCE: ................. c $B!$?""""""?,>>  A c Q  9k}?/b? >>  l "Q    232 c VTkx?}J? >>  l "V 26 c X  PNڄ?CݩvC? >>  X    D c h*)'??? >>  h A c ~ٜ&ڐ?CݩvC? >>  ~ C c ~W#4f+?Gz? >>  ~  B c }V bn$)?yEy <w  inE}V < light-harvesting chlorophyll a/b-protein complex c P  38p?˗d1? 8 >>Legend  l "@P  3 8 \L(1) %(1) \L(2) %(2) \L(3) %(3) \L(4) %(4) c "QwkG@ (>>D@YL  l "& i ( FOLDING INITIATION INDEX: c $eX@ ףp= ׿ >>BXB  l "&  SEQUENCE NUMBER c >Ft?<>>_202 &  c >Ft?<>>_202 &   R  `# R  `# R `# R ` `># R   R   R l "@l= q? R P@  R  l "? R P@2 p@ R   R   R l "@2? R P@N9 R  l "c'? R P@ N  l@ R  R  R  R  R  R  _1041328878  ac(@=xIOle ObjInfo Contents%     ; !"#$%&'()*+,-./0123456789:=>@ABCDEFGHIJKLMNOPQRCPYA 4,210 #  $@X r rr??0A1HRCX_AA.  ?@@@@@@ @"@$@&@(@*@,@.@0@1@2@3@4@5@6@7@8@9@:@;@<@=@>@?@@@@@A@A@B@B@C@C@D@D@E@E@F@F@G@G@H@H@I@I@J@J@K@K@L@L@M@M@N@N@O@O@P@@P@P@P@Q@@Q@Q@Q@R@@R@R@R@S@@S@S@S@T@@T@T@T@U@@U@U@U@V@@V@V@V@W@@W@W@W@X@@X@X@X@Y@@Y@Y@Y@Z@,t,t,t,t,t,t,t,t,t,t r !`rh??A1HRCX_B r !`rh??A1HRCX_C r !`rh??A1HRCX_D r `rr??A1HRCX_TMH r `rr??A1HRCX_BET r `rr??A1HRCX_TUR r `rr??A1HRCX_UND r `rh??A1HRCX_H.T  (\?ףp= ?Gz?\(\?\(\?RQ@Q@\(\@Gz@Gz@Q@)\(@(\@)\(@Q@{Gz@q= ףp@p= ף@(\@Gz?= ףp=?HzG??= ףp=??Gz?Q?q= ףp?ffffff?q= ףp?Q?Gz?Gz?Q?HzG?zG?Q?q= ףp??\(\?Gz?Q?(\?p= ף?ףp= ?\(\?(\?\(\?(\?333333???Q?Q?{Gz?ffffff?ףp= ?)\(@p= ף@Q@Q@@RQ@@333333@Gz@q= ףp@= ףp=@Q?(\? ףp= ?ffffff?(\?)\(??Q??Gz?Q?Q?RQ@\(\@Q@\(\@q= ףp@(\@q= ףp@ ףp= @p= ף@Q@\(\@ףp= @@Gz @ @\(\ @= ףp= @Gz@Gz@(\@Q@Q?Gz?(\?,t,t,t,t,t,t,t,t,t,t r `rr??A1HRCX_J r `rr??A1HRCX_MOMA r `rr??A1HRCX_MOMB r `rr??A1HRCX_INDA r `rr??A1HRCX_INDB r `rr??A1HRCX_MBET r ``rr??A1HRCX_DIG  ???????????????????????????,t,t,t,t,t,t,t,t,t,t r `rr??A1HRCX_OBS  ??????????????????????????????????????????????,t,t,t,t,t,t,t,t,t,t r ` `rr??A1HRCX_SC1 r `rr??A1HRCX_T.60 s Plot1$ `  ` C:\ORIGIN Fr,,s $^@4@@ȍ 2@@?@UUUUUU|l "N 9 S2? ף= ף=Pd1 c FY  b`ȿ&Y%ɿ>> L3  3 c FY , qb`ȿ&Y%ɿ>> L2  2 c FY  b`ȿ&Y%ɿ>> L1  1 c ?ulN?k8k w  l " cytochrome c (1hrc) c P  38p?˗d1? 8 >>Legend  l "@P  3 8 \L(1) %(1) \L(2) %(2) \L(3) %(3) \L(4) %(4) c Ys4r;Z5? $>>D@YL  l "&7 $ FOLDING INITIATION INDEX c W~t|K@TTTTTTԿ >>BXB  l "& SEQUENCE NUMBER R  h`# R h``# R h `# R   R   R l "@6( q? R P@  R  l "? R P@2 p@ R   R   R l "@2? R P@N$ R  l "c'? R P@ N ^@ R  R  R  R  R  R  _1041084091ac(@= + ֱOle ObjInfo ContentsCPYA 4,210 #  $@X r ``??HELSITE_A S㥫?#~j?HzG?#~j?X9v?V-?)\(?MbX9?jt?p= ף?oʡ?J +?"~?Cl?Zd;O?S?HzG? r  "!`??HELSITE_B  N''t,t,tN't,t,t,Ncap,t,t,N1,t,t,N2,t,t,N3,t,t,tN4,tN5midC5C4C3C2C1CcapC'C'' s Graph1$:^:^C:\ORIGIN,FrXXs ?1@?q@ZdS??Pl "-cS2? ף= ף=Pd1 c \ +e+h>$~ ? (>>  l "&Q ( FOLDING INITIATION FREQUENCY c D@??a?>>OR  0 c D@??a?>>OL  0 c D@??a?>>OT  0 c D@??a?>>OB  0 c U ]qyV!@'(j $>>BXB  l "&U ] $ POSITION IN THE HELIX R ``># R  l "!` R   R   R l "YW@pL? R @L R  l " R  P R   R   R l ":1@F\? R @-? R  l " R  P R  R  R  R  R  R  _1041083917ac(@=@ƍ( Ole ObjInfoContentsCPYA 4,210 #  $@X r ??Data1_A ?333333?ffffff???@@333333@@ffffff@@ @333333 @ @,t,t,t,t,t,t,t,t,t,t,t,t,t,t,t,t r `??Data1_B ??A`"?A`"?A`"?A`"?A`"?Gz?7A`?\(\?oʡ? ףp= ?^I +?#~j?,t,t,t,t,t,t,t,t,t,t,t,t,t,t,t,t r `??Data1_C M?M??B`"?V-?~jt?~jt?S?'1Z?'1Z?h|?5?S㥛?uV??,t,t,t,t,t,t,t,t,t,t,t,t,t,t,t,t s Plot1$:^:^C:\ORIGIN FrXXs ?@?@'vb'v@??@l "}@OS2? ף= ף=Pd1 c F0/aaȿɕ!ɿ>> L2  2 c Fu/aaȿɕ!ɿ>> L1  1 c " H y<ӆI?,T=?,>>  A c "  Ĺ?ͧOM?,>>  \A c V41۹^&@@Puy? X>>@YR  l "&u X + + FIS / Helix Ratio of Correct Predictions c d8p?˗d1? , >>Legend  l "@d , \L(1) %(1) \L(2) %(2) c cS]-n?Puy? \>>D@YL  l "&i \ o o Prediction Sensitivity for Folding Initiation Segments (FIS) c @Y7ʿ >>BXB  l "&   FIP threshold R `# R `# R   R   R l "@<? R P@L R  l "? R P@Z? R   R   R l "1@I? R P@? R  l "c, i ? R P@@ R  R  R  R  R  R  UR nWAʪH̺];++~u3Y dʠ nUȪAf./)qY@97@./Ck_Y{Z@f\ seПA2 =H)@ E="is+A4-WeAO4Ms 2gs5~ spDP[Ez M+/ʁ9IkAAqnoZ&. Ҵ;u!o;H-֠GyNU44Gk 2%@}xUy'P+XxPM 23:wŃŇκ`74;zV5ʹ؋IZk=yj|))9siS;ޤJk4HH sڎbi)^}Jϕ/m E70WI;Kэ45kOS_JэtJ'Hэ\2]RHi&#eFJ˛CAjnR+EW}>\~$UBBΕo8ΡiCC{rÃ<. k (뼠u!AF<)a> ԧueX--k6-6{(mnlc){DأүJ;5tCiI=4-'[J[{JZ&EP3ҫRt}{NT~iM)Ҷ?/m)EPڞi֟H=4>$%E jg_Β{(K|)Ƈ_n{(+3ėUZ!9k׵~#63,əٴm7m%gv-'9s ]Rovn8"1*͎/]-En]aI[r7,=Mz:@kK;(ck)H[~*9)njWmo?vkE '>əagH%g,#əuTo ,93n(nu8FzImԖuRə߱\rfE?O$ge4Crfms{̶`=!Ҟ3J nwl;Hwg{T-c6'JJliOR͞.͵gh˞HErf'oj[k%:iəjGnrO-zM:ީ,X:[rv|t5*ekZjv}tLꉊ{t=eq][.tnΕH,Kڽ>mi-#~{{=ET=[z@XΑf.*?k穵ٔtToWkJi=N%g~j/Q$gUK%gֵMKI|I KIμI+WHυWj}RU}3wҵ3WUKH90-Frax}+uxa $g֦}H_I|Am봧7JOI9_ϽmY*əh$+9sc%gEWrfڶ[m$g6͊/fJ.ۤ%g>-ސVcKI"əicwJK\L;+vt)ՎBGJGm/g^Iѧ}3֌H3$g^OO= '-y,KK%g jJHH<$]"9s!aTə>_ÂZ!E7`4WrffRiSəh6X&3hCə1ڰyTzI>cÒ39ZNOU{ЬYJ{|.(97'jK).93/&3[5FMrf[_IJI;9N &)U~򿤋%g.Lӓ7K+}[ic@zEr GALXrsX\ZCrf&mN,6ن{J{Hv8:Brj3$g΍P]&9sUlM6əbk=[[zTr)k3oVfII|E[+zJ[Kѭw|=i)uĝ)ugOBOGvE|}Jəkiwī;@{"d|CI?+HzSrjH_J|G['DZWrf}ͤFI8Gi֟K?WB:Nrj ,9s~_mk$gU'tݭhτKJŷI||jnv* 3NUk|]Uލ2X]Μ@KVvZufh=@+~!e-&$k`}B֯5Z`գFZjfmk ufi9g Aa*G>1rb|KVvZufh=@+~!e-&$k`}B֯5Z`գFZjfmk ufi9g Aa*GߒFZe-Њu6FY˴ 4Z~բk56ҚP3k X;h]5Kˡdx;1VWD4dzP/kVDC4ZMIV_kh(ZGG &BkC)Zfr(G@ô1Tb4u'ƿuxs-ZYiiɚ^Yh#hL@ U[jQVXiMֆR.͚P6Yic:iNWD4dzP/kVDC4ZMIV_kh(ZGG &BkC)Zfr(G@ô1Tb4u'ƿgʫrb|KVvZufh=@+~!e-&$k`}B֯5Z`գFZjfmk ufi9g Aa*GU91%Q+k;-:Y3ZQ?mih5UjZ-JQk# 5PօYʳ 0m X#M݉o`ߒFZe-Њu6FY˴ 4OB|jh(ZGG &BkC)Zfr(G@ô1Tb4u'ƿ;ϫrb|KVvZufh=@+~!e-&$k`}B֯5Z`գFZjfmk ufi9g Aa*G\U91%Q+k;-:Y3ZQ?mih5UjZ-JQk# 5PօYʳ 0m X#M݉oߒFZe-Њu6FY˴ 4XߪPk %XhքY[hm(ABݬYZYhhu6J㑦7zUNoINKN ; Y5?;1lJ٠%N[!,ft'5;䪐Ytfp ٴۢDPPY܋ЋE-x7 -tZXt?KiMKߙ#7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tń.&*M:j:h-کG.j" .RQi.jˣZJV:j-z:hn=:pv@WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg G'n麘ન66}nFꔛh3($JE ʪ9. j+Zꨡ.n&[蠞bb]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL[L8;h.WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg Gf3ܾUQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tń.&*M:j:h-کG.j" .RQi.jˣZJV:j-z:hn=:p@WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg Gt]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1聮 jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-&>u1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tu1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tż jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL[̿z~KWE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg GS/u1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tńoC t]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1-]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL[L@WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg G'6u1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=tt]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1;]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL[L^@WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg GQ(.WE5ɵQGsC4R-5E;ȝEMD%Q* 5ePVqyTP\ WG uQOpc4Qg G[.&*M:j:h-کG.j" .RQi.jˣZJV:j-z:hn=:?u1UQMmrmQ)7G umN=rgtQQpI i.jˣZJV:j-z:hn=:wo麘ન66}nFꔛh3($JE ʪ9. j+Zꨡ.n&[蠞bb*t]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1εi1jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-{t]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1񖮋 jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1[z:צM:j:h-کG.j" .RQi.jˣZJV:j-z:hn=:?zbڨ!Snۢz&(2(<*Ej⺨13nVۣz~ot]LpUTS\u>7D#uB]s[S]D\P\eGȕPYpuP[\7Fu-Jp{tPO1o1񖮋 jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)->u1UQMmrmQ)7G umN=rgtQQpIBMsUs\"WBeQCmq]SMD+uA=t<{ jjkhN9Zknv;KTjˠ*j뢞:hθ%Zn)-^bڨ!Snۢz&(2(<*Ej⺨13nVۣz~Wo麘ન66}nFꔛh3($JE ʪ9. j+Zꨡ.n&[蠞bb>-]\&F HrsP#wF5D(4AY5QA-r%TV+\5E=uDqKR7SL?n1O<_+_^u=W◾7>xxxvt}g|7 zsu!_/qx?1ŋ7~퍯֯>+O?ׯ?|uS<~?8g^ۣ߽g?^?⇇o%^yW~#ϻz}ӻ_'}=KO??;(?tߏ~iw;?z>?;?'{Kg_?>q*ߧ⯽7KڇƷ>??GO_rL{}o|矕~~o/>7>k·-7A}x-LǗ'~./տs2&KO6|?Ou1_׾Zʛo/|ͷo~+卯!\~R[r1z%Sdÿj}?}{o^?Խxo1}vү<~OqE%4^裏>O>?}?X7>-%|Fxn{O/_L5'/Q5ODd"0  # A2A׌ێlAuK+n`!A׌ێlAuK+% =~ 0T8x՚hUs=#`i2KL\q&[2!+&CbeG !3鐹ve(Z]AVVkRZglúbІ+ocMɟz=~\o(q I?%Ĕ#|RGIR Р!/nݠU~mY_-pohΧvdE`ܾo=]{amfwfσό? 2)]8HY>d5fS,X1b[W6j4v9n*.gRfUA=(qA=s4- Ow\|d\.O5PCŏ̬[ [BtCoa+*|ѷ }+ <R|?jD&AYN{k m c\X3Ǹ1֌Qb\X3ڤĸIU֌֌faㆭb\X3ƅ5݈a܍%g51|0ŸC➟Gc91h=V#ŭq'j'ƥR7 QhFqQqVkwvc\z9%q ~qi 7uF[?t 0;^;)f_byby/KǬF1<ٗ}kVR>bQ>be)fG10ϣ}mt*\,>Ӧ}9^ =|C-nN(N(KY-0n@6 1jFjڭ.06G6Ǹ^rKX-aj VK36H߳v7F[|r/;aVw~l~W>)cKwo]{Y%FyEn9G.T]<*ȓrHG~<"'N@U>jӢv5&S⸪g:?T |B}\[*oR9oJEs Ϲw8gxsr~_WF;(wȸul{nvYgb}Xs7Kz=߳ ,NS(F/ w*(b0ߵkS, xc OቆO4x͙[hUϙf/7&^(%MְQX-Ii*!U_(VK B%(B0/ FB/A|(*%36$3w9s93 %&!b:?O?,!*JFK[~#"GՔ<*'1RUhź=] $#[R|.&z@]RݽMKN3EE۽vga~N[ucom%c w̴Z~pNNk[~Di ã]([LGYBO򾨗G@]AnpnW"+иVSBIK@I+8V.z lEsb2v@mA9}F l {xQv mۥ}@ y]Zpnվ(` 7@8Wr.eйHpnUA#QV xlLDI(ۘ xuLMMie;S)38qReҠ(arO CC1dX<EpZ_1 ZWVr7K"+|Y]!x簸 {bok:B|RTaznazv³.LϏ& OR\haw׫do'}j}(213s5ژ܋{owށPRLs9sf۽$bUhauj~tWv_AP}߲з.Tx} wHG'B}Q5:w5[n"L*(at3:rgt '#.EA@NF]5j 7SLda&r 9wBS6r2nFnFANF]s 3=d1crr2n!7̐[CNF]!w9+qd܍8r7 '#X9w"܉rIdH$rϦp)&Supif"'#ܥ|<×.km~( L sN9붑1h!haDܴܴG0ׯ#ׯcMCM\\z?8'1]ꗀ;ky]Z(i|ֆ|7I/"ʒm\b#YNQKQZV/VG5oɡqlhd644^G&Ĩz',/Q5/O=~WFGGOdR9CN{ |ÁE ~DzNQĈ5Y*wCDd%0  # A2eRXW[#^cuDn`!eRXW[#^cu6 [ 5axXklsهvM Am@(ZJZ`m NeDPDZPmH8"U[խQ#TU*?ΝxKf3={>k0ɘh3fuL!` (c\f:"G~4%5~PgǻS-y'ND0~F`p^ywOovAD0){S~Os'>Mx2Ӟұ|Mk]!2pBV(=TԏW8 k!EF8Pp]`z<㮍qWxÊwML[n,S=Ԗ|-I=nJ6{]gb䪇e%n(鸛 c%#)B;܄]f/FYgrY=cjl#E ӑ:HaW%x:Rsth<)1t'ӑ^-OG []6OG gӑNV(`i,OVbMPh¾R!S˄ŀ*BkkkrWqKH` ޘ lo6jwvo l ^F~c9UTՙz;>l8pPAi8pPAi8pPAi8pPAi8pPAi8pPA*M\5;V3b/W3ڢ5\""8jXB7fZ772[ُ`zj}zӬGF 5 sO؈8lkM1D(N?. ; V{2\`+gƞ6Y\cK3K8y(_h~|G1_ }]>*zEXsKПD/Vhcn%Ck6Z8FZDW#hC'n=j΍] K3D\Y֗>Qtkq-;GɸQ[9k*&Ϩy;)9MM4m;д@ӞSҴ4MsMW9?&Ҵp%M) M]4}Z4+ M$4|4&i;2C #M73饖Kρs4vR'Mg֕TzeP_nQJ<l=kw^Oo&Rm)Rπ7l<.tyRmsΓsv^68Toשqa |(8%["# ~I*:EI|U)MǡۙogM:=HYHX&9E|8{†ÒgjҨgU^i'lh&"<6wkwh]d*L~5\\w=?ȍE1<tg5_QUGh\ ]RކN)-=CNTî`U1=qmV]O:Q\PH̏64y^֞~ZvݝgaП^2:`5AK,;r${?K䩁l<ɪ%(wLC _La>1Table?'SummaryInformation(DocumentSummaryInformation8CompObjj Oh+'0 0 HT p |  MPrediction of initiation sites for protein folding with α-helix preferencesfredZAVOD ZA FIZIKUAVONormalAZAVOD ZA FIZIKU4VOMicrosoft Word 8.0a@2~@dR@܄]@jY ՜.+,D՜.+,< hp   PMF SPLITn E7  MPrediction of initiation sites for protein folding with α-helix preferences Title 6> _PID_GUIDAN{C1FD1440-EDE1-11D4-8BD3-08002B96244F}  FMicrosoft Word Document MSWordDocWord.Document.89q [D@D Normal1$CJOJQJ_HhmH sH tH P@P Heading 1 $1$@& 5;CJOJQJhmHsHuD@D Heading 2$$dh@&a$ 5OJQJ<A@< Default Paragraph Font4&@4 Footnote Reference,@, Header  9r , @, Footer  9r &)@!& Page Number(U@1( Hyperlink>*B*>Z@B> Plain Text1$CJOJQJhu8YR8 Document Map-D OJQJ4B@b4 Body Text$a$OJQJ<Yv  <YvyP-1cAttO]808]88889D9q9                "     P&(-18>lB>KiSVXcj~kmAtu~tO]y, y S='x ')hy )u-4?M,mVuvDv_v|Y Pָehz|}~3iy98h{ '*!!t79<TVYqsy::::  O2$eRXW[#^cu92$ dXx9A2$A׌ێlAuK+F2$'Ŭ~r%`2{6N@" (  \  S A ?"\  S A ?"\   S  A ?""\   S  A ?"B S  ?>Vj~kf !4 T%4#4  j#u42)8)))P*V*****!+$+SSSTTTT7T9TBTCTLTMTYTZTfThTkTmTpTrTuTwTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTU-UUUUUUVgVuVXX9Z@Zvczcddddjjwx)x0xxx]ykyyyyyyyi}r}}}}}9~?~f~q~?KHLMRX]^ftz68„  ׆ކ -1?BMQ_b "09ЎӎːҐőS] 2:;BCGוĖȖ֖ٖ`gəЙryip $v;BZ^_OR%A!pz& @IRS[!,_jY { ^!i!!!["f"""#$#[#e#$'$$$$$%%&&['e'(q(((**++, ,H,R,,,,,-$-`-j-----3.<.y....//K/T/////0&0c0l00000uAACCEFAIPI RiRsSSSSSSSTTTT6T9TATCTKTMTXTZTeThTjTmToTrTtTwTTTTTTTTTTTTTTTTTTTTTTTTTTTTKUXUVW7W?W XXYY9Z@ZZZiijjoktkklll6m9mooss̃׃EMFS .=ˈu0!AJU[Ս܍t>Eܑ!(ʒgl!2HIRWg)6 ̚~6AOZݟǠOWڡZ_!%:BEPAC^}:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::ZAVOD ZA FIZIKU1C:\WINDOWS\TEMP\AutoRecovery save of PERBIO28.asdZAVOD ZA FIZIKU1C:\WINDOWS\TEMP\AutoRecovery save of PERBIO28.asdZAVOD ZA FIZIKU1C:\WINDOWS\TEMP\AutoRecovery save of PERBIO28.asdZAVOD ZA FIZIKU1C:\WINDOWS\TEMP\AutoRecovery save of PERBIO28.asdZAVOD ZA FIZIKU#C:\My Documents\perbio\PERBIO28.docZAVOD ZA FIZIKU#C:\My Documents\perbio\PERBIO28.docZAVOD ZA FIZIKUA:\PERBIO28.docZAVOD ZA FIZIKU A:\PERB.docLuka A:\PERB.docZAVOD ZA FIZIKU A:\PERB.doc@HP LaserJet 6L PCLLPT1:PCL5EMS3HP LaserJet 6L PCLHP LaserJet 6L PCL@g XX @MSUDHP LaserJet 6L PCL&d HP LaserJet 6L PCL@g XX @MSUDHP LaserJet 6L PCL&d DD4$dDD $$$($)$1$3$9$;>?EGIKLNPsSsTsesfsjsks~s?? QQQQQQQQ@@@@@@@ @@@0@@@@@@*@X@@.@`@@:@x@@D@@@N@@@V@@@^@b@@@h@@@l@@@t@@@@@@@ @@@L@@@T@@@\@@@d@@@@p@@@x@@@@@@@@@@@@G:Times New Roman5Symbol3& :ArialE5  Lucida ConsoleA& Arial Narrow?5 :Courier New5& :Tahoma"A hQQ٢9&YE " d7 VKPrediction of initiation sites for protein folding with -helix preferencesZAVOD ZA FIZIKUZAVOD ZA FIZIKU