Predicting the behaviour of proteins in hydrophobic interaction chromatography. 1: Using the hydrophobic imbalance (HI) to describe their surface amino acid distribution

This paper focuses on the prediction of the dimensionless retention time of proteins (DRT) in hydrophobic interaction chromatography (HIC) by means of mathematical models based on characteristics of the surface hydrophobicity distribution. We introduce a new parameter, called hydrophobic imbalance (HI), obtained from the three-dimensional structure of proteins. This parameter quantifies the displacement of the superficial geometric centre of the protein when the effect of the hydrophobicity of each amino acid is considered. This parameter is simpler and less expensive than those reported previously. We use HI as a way to incorporate information about the surface hydrophobicity distribution in order to improve the prediction of DRT. We tested the performance of our DRT predictive models in a set of 15 proteins. This set includes four proteins whose DRTs are known as very difficult to predict. By means of the variable HI, it was possible to improve the predictive characteristics obtained by models based on the average surface hydrophobicity (ASH) by 9.1%. Also, we studied linear multivariable models based on characteristics determined from the HI. By using this multivariable model, a correlation coefficient of 0.899 was obtained. With this model, we managed to improve the predictive characteristics shown by previous models based on ASH by 31.8%.     

Affinity partitioning of human antibodies in aqueous two-phase systems

The partitioning of human immunoglobulin (IgG) in a polymer-polymer and polymer-salt aqueous two-phase system (ATPS) in the presence of several functionalised polyethylene glycols (PEGs) was studied. As a first approach, the partition studies were performed with pure IgG using systems in which the target protein remained in the bottom phase when the non-functionalised systems were tested. The effect of increasing functionalised PEG concentration and the type of ligand were studied. Afterwards, selectivity studies were performed with the most successful ligands first by using systems containing pure proteins and an artificial mixture of proteins and, subsequently, with systems containing a Chinese hamster ovary (CHO) cells supernatant. The PEG/phosphate ATPS was not suitable for the affinity partitioning of IgG. In the PEG/dextran ATPS, the diglutaric acid functionalised PEGs (PEG-COOH) displayed great affinity to IgG, and all IgG could be recovered in the top phase when 20% (w/w) of PEG 150-COOH and 40% (w/w) PEG 3350-COOH were used. The selectivity of these functionalised PEGs was evaluated using an artificial mixture of proteins, and PEG 3350-COOH did not show affinity to IgG in the presence of typical serum proteins such as human serum albumin and myoglobin, while in systems with PEG 150-COOH, IgG could be recovered with a yield of 91%. The best purification of IgG from the CHO cells supernatant was then achieved in a PEG/dextran ATPS in the presence of PEG 150-COOH with a recovery yield of 93%, a purification factor of 1.9 and a selectivity to IgG of 11. When this functionalised PEG was added to the ATPS, a 60-fold increase in selectivity was observed when compared to the non-functionalised systems.     

Is it possible to predict the average surface hydrophobicity of a protein using only its amino acid composition?

Hydrophobicity is one of the most important physicochemical properties of proteins. Moreover, it plays a fundamental role in hydrophobic interaction chromatography, a separation technique that, at present time, is used in most industrial processes for protein purification as well as in laboratory scale applications. Although there are many ways of assessing the hydrophobicity value of a protein, recently, it has been shown that the average surface hydrophobicity (ASH) is an important tool in the area of protein separation and purification particularly in protein chromatography. The ASH is calculated based on the hydrophobic characteristics of each class of amino acid present on the protein surface. The hydrophobic characteristics of the amino acids are determined by a scale of aminoacidic hydrophobicity. In this work, the scales of Cowan-Whittaker and Berggren were studied. However, to calculate the ASH, it is necessary to have the three-dimensional protein structure. Frequently this data does not exist, and the only information available is the amino acid sequence. In these cases it would be desirable to estimate the ASH based only on properties extracted from the protein sequence. It was found that it is possible to predict the ASH from a protein to an acceptable level for many practical applications (correlation coefficient > 0.8) using only the aminoacidic composition. Two predictive tools were built: one based on a simple linear model and the other on a neural network. Both tools were constructed starting from the analysis of a set of 1982 non-redundant proteins. The linear model was able to predict the ASH for an independent subset with a correlation coefficient of 0.769 for the case of Cowan-Whittaker and 0.803 for the case of Berggren. On the other hand, the neural model improved the results shown by the linear model obtaining correlation coefficients of 0.831 and 0.836, respectively. The neural model was somewhat more robust than the linear model particularly as it gave similar correlation coefficients for both hydrophobicity scales tested, moreover, the observed variabilities did not overcome 6.1% of the mean square error. Finally, we tested our models in a set of nine proteins with known retention time in hydrophobic interaction chromatography. We found that both models can predict this retention time with correlation coefficients only slightly inferior (11.5% and 5.5% for the linear and the neural network models, respectively) than models that use the information about the three-dimensional structure of proteins.     

Aqueous two-phase systems for protein separation: a perspective

Aqueous two-phase systems (ATPS) that are formed by mixing a polymer (usually polyethylene glycol, PEG) and a salt (e.g. phosphate, sulphate or citrate) or two polymers and water can be effectively used for the separation and purification of proteins. The partitioning between both phases is dependent on the surface properties of the proteins and on the properties of the two phase system. The mechanism of partitioning is complex and not very easy to predict but, as this review paper shows, some very clear trends can be established. Hydrophobicity is the main determinant in the partitioning of proteins and can be measured in many different ways. The two methods that are more attractive, depending on the ATPS used (PEG/salt, PEG/polymer), are those that consider the 3-D structure and the hydrophobicity of AA on the surface and the one based on precipitation with ammonium sulphate (parameter 1/m*). The effect of charge has a relatively small effect on the partitioning of proteins in PEG/salt systems but is more important in PEG/dextran systems. Protein concentration has an important effect on the partitioning of proteins in ATPS. This depends on the higher levels of solubility of the protein in each of the phases and hence the partitioning observed at low protein concentrations can be very different to that observed at high concentrations. In virtually all cases the partition coefficient is constant at low protein concentration (true partitioning) and changes to a different constant value at a high overall protein concentration. Furthermore, true partitioning behavior, which is independent of the protein concentration, only occurs at relatively low protein concentration. As the concentration of a protein exceeds relatively low values, precipitation at the interface and in suspension can be observed. This protein precipitate is in equilibrium with the protein solubilized in each of the phases. Regarding the effect of protein molecular weight, no clear trend of the effect on partitioning has been found, apart from PEG/dextran systems where proteins with higher molecular weights partitioned more readily to the bottom phase. Bioaffinity has been shown in many cases to have an important effect on the partitioning of proteins. The practical application of ATPS has been demonstrated in many cases including a number of industrial applications with excellent levels of purity and yield. This separation and purification has also been successfully used for the separation of virus and virus-like particles.     

Aqueous two-phase partitioning of milk proteins

Milk of transgenic pigs secreting recombinant human Protein C (rHPC) was used as a model system to determine the utility of aqueous two-phase extraction systems (ATPS) for the initial step in the purification of proteins from milk. The major challenges in purification of recombinant proteins from milk are removal of casein micelles (that foul processing equipment) and elimination of the host milk proteins from the final product. When milk was partitioned in ATPS composed of polyethylene glycol (PEG) and ammonium sulfate (AS), the phases were clarified and most of the caseins precipitated at the interphase. The partition coefficients of the major milk proteins and rHPC were dependent upon the molecular weight of the PEG used in the ATPS. Higher-partition coefficients of the major whey proteins, Β-lactoglobulin, and α-lactalbumin were observed in ATPS made up of lower molecular-weight PEG (1000 or 1450) as compared to systems using higher molecular-weight PEG. Lowering the pH of the ATPS from 7.5 to 6.0 resulted in increased precipitation of the caseins and decreased their concentration in both phases. rHPC had a partition coefficient of 0.04 in a system composed of AS and PEG 1450. The rHPC in pig milk was shown to be highly heterogenous by two-dimensional gel electrophoresis. The heterogeneity was owing to inefficient proteolytic processing of the single chain to the heterodimeric form and differences in glycosylation and other post-translational processing. Differential partitioning of the multiple forms of purified rHPC in the ATPS was not observed. rHPC after processing in ATPS was recovered in a clear phase free of most major milk proteins. ATPS are useful as the initial processing step in the purification of recombinant proteins from milk because clarification and enrichment is combined in a single step.     

Application of central composite design to the optimisation of aqueous two-phase extraction of human antibodies

The partition of human antibodies in aqueous two-phase systems (ATPSs) of polyethylene glycol (PEG) and phosphate was systematically studied using first pure proteins systems and then an artificial mixture of proteins containing 1mg/ml human immunoglobulin G (IgG), 10mg/ml serum albumin and 2mg/ml myoglobin. Preliminary results obtained using pure proteins systems indicated that the PEG molecular weight and concentration, the pH value and the salts concentration had a pronounced effect on the partitioning behaviour of all proteins. For high ionic strengths and pH values higher than the isoelectric point (pI) of the contaminant proteins, IgG could be selectively recovered on the top phase. According to these results, a face centred composite design was performed in order to optimise the purification of IgG from the mixture of proteins. The optimal conditions for the isolation of IgG were observed for high concentrations of NaCl and low concentrations of both phase forming components. The best purification was achieved using an ATPS containing 8% (w/w) PEG 3350, 10% (w/w) phosphate pH 6 and 15% (w/w) NaCl. A recovery yield of 101+/-7%, a purity of 99+/-0% and a yield of native IgG of 97+/-4% were obtained. Back extraction studies of IgG to a new phosphate phase were performed and higher yields were obtained using 10% phosphate buffer at pH 6. The total extraction yield was 76% and the purity 100%.     

Selective separation and enrichment of proteins in aqueous two-phase extraction system

A simple aqueous two-phase extraction system(ATPS) of PEG/phosphate was proposed for selective separation and enrichment of proteins.The combination of ATPE with HPLC was applied to identify the partition of proteins in two phases.Five proteins (bovine serum albumin,Cytochrome C,lysozyme,myoglobin,and trypsin) were used as model proteins to study the effect of phosphate concentration and pH on proteins partition.The PEG/phosphate system was firstly applied to real human saliva and plasma samples,some pro...     

Effect of salt additives on partition of nonionic solutes in aqueous PEG-sodium sulfate two-phase system

Partition of 12 nonionic organic compounds in aqueous PEG-8000-Na(2)SO(4) two-phase system was examined. Effects of four salt additives (NaCl, NaSCN, NaClO(4), and NaH(2)PO(4)) in the concentration range from 0.027 up to ca. 1.9 M on binodal curve of PEG-sulfate two-phase system and solute partitioning were explored. It was found that different salt additives at the relatively high concentrations display different effects on both phase separation and partition of various nonionic solutes. Analysis of the results indicates that the PEG-Na(2)SO(4) ATPS with the up to 0.215 M NaCl concentration may be viewed as similar to the ATPS without NaCl in terms of the Collander equation's predictive ability of the partitioning behavior of nonionic compounds. All ATPS with each of the salt additive used at the concentration of 0.027 M may be viewed as similar to each other as the Collander equation holds for partition coefficients of nonionic solutes in these ATPS. Collander equation is valid also for the compounds examined in the ATPS with additives of NaSCN and NaClO(4) at the concentrations up to 0.215 M. The observed similarity between these ATPS might be explained by the similar effects of these two salts on the water structure. At concentrations of the salt additives exceeding the aforementioned values, different effects of salt additives on partitioning of various nonionic solutes are displayed. In order to explain these effects of salt additives it is necessary to examine the intensities of different solute-solvent interactions in these ATPS within the framework of the so-called Linear Solvation Energy Relationship (LSER) model.     

双水相萃取结合液相色谱法分离蛋白质

建立了PEG/( NH4)2SO4双水相体系萃取富集,结合液相色谱分离分析多种蛋白质的方法.考察了无机盐种类和浓度、PEG分子量、pH值和温度等因素对双水相形成以及对细胞色素C、肌红蛋白、牛血清白蛋白、溶菌酶、胰蛋白酶分配行为的影响.结果表明,上述5种蛋白在室温、pH 3.5～9.0范围内,可在15％ PEG-4000/10％ (NH4)2SO4双水相体系中得到富集,且主要集中在下相.同样条件下,血清中的高丰度蛋白在上下相均有分配,下相分配量较大.通过双水相萃取分离蛋白质及对液相色谱一定时间段的色谱峰收集,可初步实现血清中高丰度蛋白质的分离去除.     

Direct purification of pectinase from mango (Mangifera Indica Cv. Chokanan) peel using a PEG/salt-based Aqueous Two Phase System

An Aqueous Two-Phase System (ATPS) was employed for the first time for the separation and purification of pectinase from mango (Mangifera Indica Cv. Chokanan) peel. The effects of different parameters such as molecular weight of the polymer (polyethylene glycol, 2,000-10,000), potassium phosphate composition (12-20%, w/w), system pH (6-9), and addition of different concentrations of neutral salts (0-8%, w/w) on partition behavior of pectinase were investigated. The partition coefficient of the enzyme was decreased by increasing the PEG molecular weight. Additionally, the phase composition showed a significant effect on purification factor and yield of the enzyme. Optimum conditions for purification of pectinase from mango peel were achieved in a 14% PEG 4000-14% potassium phosphate system using 3% (w/w) NaCl addition at pH 7.0. Based on this system, the purification factor of pectinase was increased to 13.2 with a high yield of (97.6%). Thus, this study proves that ATPS can be an inexpensive and effective method for partitioning of pectinase from mango peel.     

Partitioning of lactate dehydrogenase from bovine heart crude extract by polyethylene glycol-citrate aqueous two-phase systems

Aqueous two-phase systems (ATPS) composed of polyethylene glycol (PEG)-citrate have been used for enzyme partitioning studies. The behavior of lactate dehydrogenase (LDH) from bovine heart crude extract was analyzed using a two-level factorial design in which the PEG molar mass and concentration, the citrate concentration were selected as independent variables, while the purification factor, the partition coefficient (K) and the activity yield were selected as responses. The statistical analysis revealed the effect of PEG molar mass on K. LDH exhibited a better partitioning toward PEG-rich phase and the highest K value (1079.81) was obtained with 42% (w/w) PEG 400 and 7.5% (w/w) citrate concentration. PEG molar mass also influenced the purification factor of the enzyme in the top phase. Possibly these ATPS remove inhibitors present in the extract affording higher enzyme yield.     

Affinity partitioning of plasmid DNA with a zinc finger protein

The affinity isolation of pre-purified plasmid DNA (pDNA) from model buffer solutions using native and poly(ethylene glycol) (PEG) derivatized zinc finger-GST (Glutathione-S-Transferase) fusion protein was examined in PEG-dextran (DEX) aqueous two-phase systems (ATPSs). In the absence of pDNA, partitioning of unbound PEGylated fusion protein into the PEG-rich phase was confirmed with 97.5% of the PEGylated fusion protein being detected in the PEG phase of a PEG 600-DEX 40 ATPS. This represents a 1322-fold increase in the protein partition coefficient in comparison to the non-PEGylated protein (Kc = 0.013). In the presence of pDNA containing a specific oligonucleotide recognition sequence, the zinc finger moiety of the PEGylated fusion protein bound to the plasmid and steered the complex to the PEG-rich phase. An increase in the proportion of pDNA that partitioned to the PEG-rich phase was observed as the concentration of PEGylated fusion protein was increased. Partitioning of the bound complex occurred to such an extent that no DNA was detected by the picogreen assay in the dextran phase. It was also possible to partition pDNA using a non-PEGylated (native) zinc finger-GST fusion protein in a PEG 1000-DEX 500 ATPS. In this case the native ligand accumulated mainly in the PEG phase. These results indicate good prospects for the design of new plasmid DNA purification methods using fusion proteins as affinity ligands.     

Xylanase recovery effect of extraction conditions on the aqueous two-phase system using experimental design

The partitioning of xylanase produced byPenicillium janthinellum in aqueous two-phase systems (ATPS) using poly(ethylene glycol) (PEG) and phosphate (K₂HPO₄/KH₂PO₄) was studied employing a statistical experimental design. The aim was to identify the key factors governing xylanase partitioning. The interactions of five factors (PEG concentration molecular weight, concentration of buffer K₂HPO₄/KH₂PO₄, pH, and NaCl concentration) and their main effects on the partition coefficient (K) were evaluated by means of a 2⁵ full-factorial experimental design with four center points. The %PEG, %NaCl, and pH were the most important factors affecting the response variable (K). Response surface methodology (RSM) was adopted and an empirical second-order polynomial model was constructed on the basis of the results. The optimum partition conditions were pH 7.0, PEG = 8.83% and NaCl = 6.02%. Adequacy of the model for predicting optimum response value was tested under these conditions. The experimental xylanase partition coefficient (K) was 2.21, whereas its value predicted by the model was 2.33. These results indicate that the predicted model was adequate for the process. PEG molecular weight and phosphate concentration did not affect the xylanase partition coefficient.     

System Establishment of ATPS for One-Step Purification of Glutamate Decarboxylase from <Emphasis Type="Italic">E. coli</Emphasis> After Cell Disruption

The partition of glutamate decarboxylase (GAD) from Escherichia coli in polyethylene glycol (PEG) and sodium sulfate aqueous two-phase systems (ATPS) has been explored with the purpose of establishing a phase system for the purification of GAD after cell disruption. The results showed that the partitioning of GAD was slightly influenced by PEG molecular weight (MW) but depended on the tie line length (TLL) and NaCl and loading sample concentrations. The optimum system obtained for GAD purification was composed of a PEG MW of 4,000, TLL of 63.5%, a volume ratio of 2.31, a loading sample concentration of 0.4 g/mL, which produced a GAD recovery of 90% with the purification fold of 73. Furthermore, the feasibility of directly purifying GAD from the cell disrupts using ATPS was evaluated. The established ATPS for GAD purification exhibited an efficient integrated purification process compared to the reported purification process in terms of purification efficiency and recovery.     

Synthesis of a Novel pH-Sensitive Methacrylate Amphiphilic Polymer and Its Primary Application in Aqueous Two-phase Systems

In this study, a novel pH-sensitive and reversible water-soluble polymer(P(ABC)) forming aqueous two-phase systems(ATPS) was synthesized by using 2-(dimethylamino)ethyl methacrylate, t-butyl methacrylate, and methyl methacrylate as monomers and 2,2'-azo-bis-isobutyronitrile as initiator. The P(ABC) could be recovered by adjusting isoelectric point (PI) to 8.4, and recovery at PI could reach 95%. ATPS was formed by 5% (w/w) P(ABC) and 10% (w/w) PEG20000. The partition coefficient K of lysozyme was 6.8, and the partition coefficient K of bovine serum albumin could reach 12.5 in the ATPS.     

Partitioning and assembly of metal particles and their bioconjugates in aqueous two-phase systems

The behavior of metal nanospheres and nanowires and their bioconjugates in aqueous two-phase systems (ATPS) is described. The ATPS used in this work comprised poly(ethylene glycol) (PEG), dextran, and water or aqueous buffer. Au and Ag nanospheres less than 100 nm in diameter partition between the PEG-rich and dextran-rich phases on the basis of their surface chemistry and can be separated on this basis. Larger Au nanospheres and wires accumulate at the interface between the two aqueous phases. The influence of polymer molecular weight and concentration on interfacial assembly of Au wires is described. DNA-derivatized nanowires at the aqueous/aqueous interface retain the ability to selectively bind to fluorescent complementary DNA. In addition, Au nanoparticles have been bound to Au wires via selective DNA hybridization at the ATPS interface. Transmission electron microscopy and thermal denaturation experiments confirm that DNA-driven assembly is responsible for the formation of the nanosphere/wire assemblies. These results demonstrate the biocompatibility of the two-phase interface and point to future use as scaffolding in biorecognition-driven assembly.     

Evaluation of a PEG/hydroxypropyl starch aqueous two‐phase system for the separation of monoclonal antibodies from cell culture supernatant

In this study, an aqueous two‐phase system (ATPS) with PEG and hydroxypropyl starch (HPS) was used to separate monoclonal antibody (mAb) from Chinese hamster ovary cell culture supernatant. The phase diagram of the PEG/HPS ATPS was determined, and the effects of NaCl addition were investigated. The results showed that NaCl addition could lead to a shift of the binodal curve and that phase separation would occur at higher PEG and HPS concentrations. The effects of NaCl addition, pH, and the load of cell supernatant on the partitioning of mAb in a PEG/HPS ATPS were investigated. It was found that with 6% cell supernatant and 15% NaCl addition at pH 6.0, the yield of mAb in the upper phase was 96.7% with a purity of 96.0%. The back‐extraction of mAb with a PEG/phosphate ATPS were also studied, and the results showed that after the two‐step extraction with ATPSs the purity of mAb could reach 97.6 ± 0.5% with a yield of 86.8 ± 1.0%, which was comparable to the purification with Protein A chromatography. These results indicate that the two‐step extraction with PEG/HPS and PEG/phosphate ATPSs might be a promising alternative for the separation of mAb from cell culture supernatant.     

Purification of plasmid DNA vectors by aqueous two-phase extraction and hydrophobic interaction chromatography

The current study explores the possibility of using a polyethyleneglycol(PEG)-ammonium sulphate aqueous two-phase system (ATPS) as an early step in a process for the purification of a model 6.1 kbp plasmid DNA (pDNA) vector. Neutralised alkaline lysates were fed directly to ATPS. Conditions were selected to direct pDNA towards the salt-rich bottom phase, so that this stream could be subsequently processed by hydrophobic interaction chromatography (HIC). Screening of the best conditions for ATPS extraction was performed using three PEG molecular weights (300, 400 and 600) and varying the tie-line length, phase volume ratio and lysate load. For a 20% (w/w) lysate load, the best results were obtained with PEG 600 using the shortest tie-line (38.16%, w/w). By further manipulating the system composition along this tie-line in order to obtain a top/bottom phase volume ratio of 9.3 (35%, w/w PEG 600, 6%, w/w NH4)2 SO4), it was possible to recover 100% of pDNA in the bottom phase with a three-fold increase in concentration. Further increase in the lysate load up to 40% (w/w) with this system resulted in a eight-fold increase in pDNA concentration, but with a yield loss of 15%. The ATPS extraction was integrated with HIC and the overall process compared with a previously defined process that uses sequential precipitations with iso-propanol and ammonium sulphate prior to HIC. Although the final yield is lower in the ATPS-based process the purity grade of the final pDNA product is higher. This shows that it is possible to substitute the time-consuming two-step precipitation procedure by a simple ATPS extraction.     

Solvatochromic relationship: Prediction of distribution of ionic solutes in aqueous two-phase systems

Partition ratios of several ionic compounds in 20 different polymer/polymer aqueous two-phase systems (ATPS) containing 0.15 M NaCl in 0.01 M phosphate buffer, pH 7.4, were determined. The differences between the electrostatic properties of the phases in all the ATPS were estimated from partitioning of the homologous series of dinitrophenylated-amino acids. Also the solvatochromic solvent parameters characterizing the solvent dipolarity/polarizability (π*), solvent hydrogen-bond donor acidity (α), and solvent hydrogen-bond acceptor basicity (β) of aqueous media were measured in the coexisting phases of the ATPS. The solute-specific coefficients for the compounds examined were determined by the multiple linear regression analysis using the modified linear solvation energy relationship equation. The minimal number of ATPS necessary for determination of the coefficients was established and 10 ATPS were selected as a reference ATPS set. The solute-specific coefficients values obtained with this reference set of ATPS were used to predict the partition ratios for the compounds in 10 ATPS not included in the reference set. The predicted partition ratios values were compared to those determined experimentally and found to be in good agreement. It is concluded that the presented model of solute–solvent interactions as the driving force for solute partitioning in polymer/polymer ATPS describes experimental observations with 90–95% accuracy.     

Prediction of retention times of proteins in hydrophobic interaction chromatography using only their amino acid composition

This paper focuses on the prediction of the dimensionless retention time of proteins (DRT) in hydrophobic interaction chromatography (HIC) by means of mathematical models based, essentially, only on aminoacidic composition. The results show that such prediction is indeed possible. Our main contribution was the design of models that predict the DRT using the minimal information concerning a protein: its aminoacidic composition. The performance is similar to that observed in models that use much more sophisticated information such as the three-dimensional structure of proteins. Three models that, in addition to the amino acid composition, use different assumptions about the amino acids tendency to be exposed to the solvent, were evaluated in 12 proteins with known experimental DRT. In all the cases analyzed, the model that obtained the best results was the one based on a linear estimation of the aminoacidic surface composition. The models were adjusted using a collection of 74 vectors of aminoacidic properties plus a set of 6388 vectors derived from these using two mathematical tools: k-means and self-organizing maps (SOM) algorithms. The best vector was generated by the SOM algorithm and was interpreted as a hydrophobicity scale based partly on the tendency of the amino acids to be hidden in proteins. The prediction error (MSE(JK)) obtained by this model was almost 35% smaller than that obtained by the model that supposes that all the amino acids are completely exposed and 40% smaller than that obtained by the model that uses a simple correction factor considering the general tendency of each amino acid to be exposed to the solvent. In fact, the performance of the best model based on the aminoacidic composition was 5% better than that observed in the model based on the three-dimensional structure of proteins.