Top PDF Phage-Host Interaction in Nature

Phage-Host Interaction in Nature

Phage-Host Interaction in Nature

is capable of making reasonable order of magnitude predictions and that can qualitatively explain the observed trends. Predator-prey models for host-virus interaction have earlier been examined by Campbell [5], Levin et al. [6], Lenski [7], Beretta et al. [8] and Thingstad et al. [9,10]. However, in these models the biophysical process of virus transport, which governs the contact rates between viruses and bacteria, was not considered. Stent [11] and Murray et al. [12] considered transport processes of viruses in aqueous environments but not in the context of a predator-prey model in an ecological setting. Our starting point is a simple toy model that incorporates virus transport within the context of a predator-prey model. We begin by examining the case of a particular isolated phage-host system with the goal of identifying the key variables that govern this system. We then extend our model to the community scale by hypothesizing the simplest evolutionary scenario that there is no selection pressure on bacterial radii, i.e., a priori, all bacterial radii are equally probable. We derive basic relations for the total concentration of bacteria and their viruses in the environment, and a basic relation for the total prokaryotic mass in the environment. Based on these results we explore questions such as, what are the critical parameters governing the system and how do variables scale with respect to these parameters? What determines the virus-to-bacterium ratio? What determines the number of species in a given environment? In what volume of water should we find this diversity? What are the bounds on the total diversity of species in Earth’s oceans? Where possible we compare our predictions to observations and conclude with suggestions for experiments to further test our model.
Show more

279 Read more

Structure and Functional Analysis of the Host Recognition Device of Lactococcal Phage Tuc2009

Structure and Functional Analysis of the Host Recognition Device of Lactococcal Phage Tuc2009

B acterial viruses (bacteriophages, or phages) of the order Cau- dovirales possess a tail that recognizes the host and ensures genome delivery upon infection. This host recognition event is mediated through binding of the tip of the tail to either a protein receptor or a carbohydrate moiety located in or on the cell enve- lope (1–3). Well-characterized examples of bacteriophages with protein receptors include coliphages lambda and T5, which rec- ognize LamB and FhuA, respectively, both located on the surface of the Escherichia coli cell envelope (4–7), and Bacillus phage SPP1, which recognizes YueB at the cell surface (8, 9). An alternative cell binding approach by prototype bacteriophage T4 follows a two- step process whereby the T4 long tail fibers first reversibly bind lipopolysaccharide (LPS) or OmpC (10–13), causing a conforma- tional change in the baseplate, which then allows irreversible bind- ing of the short tail fibers of the baseplate to LPS (14). In the case of carbohydrate-dependent host recognition, phages appear to employ a so-called baseplate, a large heteropolymeric protein- aceous organelle, in order to ensure efficient and host-specific binding (15, 16). The study of saccharidic phage receptors is still in its infancy but differs from the protein receptor model in that, due to the relative weakness of an individual carbohydrate-protein interaction, phage binding to a host is consolidated typically by a large number of receptor binding proteins (RBPs) in the phage baseplate (17, 18). The baseplate structures have only relatively recently been recognized as host recognition devices and are of particular interest among phages of Lactococcus lactis.
Show more

12 Read more

Arbovirus Structure and Interaction with Host Cells.

Arbovirus Structure and Interaction with Host Cells.

29 Analysis of the hydrophobicity of the E1 and E2 TMDs by helical wheel analysis in Heliquest and Protean shows that E1 has a more hydrophobic nature than E2. These predictions suggest that E1 and E2 are not typical amphipathic helices with clear hydrophobic and hydrophilic surfaces. Our previous genetic analysis of the effects of amino acid deletions in E2 which affected virus titer are found in the more hydrophobic face of the helix with the exception of A385 (shown in Fig. 1.7) which is in agreement with other genetic studies of SIN and RR chimeras. When SINV E1 was investigated by deletion analysis, mutations affecting titer also mapped primarily to a single face of E1 but these were found to cluster toward the more hydrophilic face. This was also unexpected because it was expected that the two adjacent faces of E1 and E2 would interact. However, this was also the case in the E1 modeling of the reconstructions of SINV and VEEV referred to above. Another feature of these sequences is that SINV E2 TM18 deletes 8 consecutive amino acids but retains full infectivity. This is also interesting because E1 and E2 are modeled as coiled coils. Because these domains are in the membrane bilayer we will assume, for this discussion, that each face of these helices is defined by amino acids within 180 ° of the first amino acid in the hydrophobic face calculated by Heliquest (Figures 1.3.1A and 1.5A). The sequence of SINV TM18 365 VYTIL 370 AVASA 375 TVAMM 380 IGVTV 385 AVLCA 390 C (underlined sequence deleted) the helix sequence does not come back into the original register until A389 (Fig. 1.8) This implies that amino acids known to affect titer, with the exception of I380, (replaces M379) relocate to the more hydrophilic face (Figure 1.3.1A).
Show more

123 Read more

Global Transcriptional Analysis of Virus-Host Interactions between Phage ϕ29 and Bacillus subtilis

Global Transcriptional Analysis of Virus-Host Interactions between Phage ϕ29 and Bacillus subtilis

Bertero MG, Bessieres P, Bolotin A, Borchert S, Borriss R, Boursier L, Brans A, Braun M, Brignell SC, Bron S, Brouillet S, Bruschi CV, Caldwell B, Capuano V, Carter NM, Choi SK, Cordani JJ, Connerton IF, Cummings NJ, Daniel RA, Denziot F, Devine KM, Dusterhoft A, Ehrlich SD, Emmerson PT, Entian KD, Errington J, Fabret C, Ferrari E, Foulger D, Fritz C, Fujita M, Fujita Y, Fuma S, Galizzi A, Galleron N, Ghim SY, Glaser P, Goffeau A, Golightly EJ, Grandi G, Guiseppi G, Guy BJ, Haga K, Haiech J, Harwood CR, Henaut A, Hilbert H, Holsappel S, Hosono S, Hullo MF, Itaya M, Jones L, Joris B, Karamata D, Kasahara Y, Klaerr-Blanchard M, Klein C, Kobayashi Y, Koetter P, Koningstein G, Krogh S, Kumano M, Kurita K, Lapidus A, Lardinois S, Lauber J, Lazarevic V, Lee SM, Levine A, Liu H, Masuda S, Mauel C, Medigue C, Medina N, Mellado RP, Mizuno M, Moestl D, Nakai S, Noback M, Noone D, O’Reilly M, Ogawa K, Ogiwara A, Oudega B, Park SH, Parro V, Pohl TM, Portelle D, Porwollik S, Prescott AM, Presecan E, Pujic P, Purnelle B, Rapoport G, Rey M, Reynolds S, Rieger M, Rivolta C, Rocha E, Roche B, Rose M, Sadaie Y, Sato T, Scanlan E, Schleich S, Schroeter R, Scoffone F, Sekiguchi J, Sekowska A, Seror SJ, Serror P, Shin BS, Soldo B, Sorokin A, Tacconi E, Takagi T, Takahashi H, Takemaru K, Takeuchi M, Tamakoshi A, Tanaka T, Terpstra P, Togoni A, Tosato V, Uchiyama S, Vandebol M, Vannier F, Vassarotti A, Viari A, Wambutt R, Wedler H, Weitzenegger T, Winters P, Wipat A, Yamamoto H, Yamane K, Yasumoto K, Yata K, Yoshida K, Yoshikawa HF, Zumstein E, Yoshikawa H, Danchin A. 1997. The complete genome sequence of the gram-positive bacterium Bacillus subtilis. Nature 390:249 –256. http://dx .doi.org/10.1038/36786.
Show more

12 Read more

Phage-Mediated Selection on Microbiota of a Long-Lived Host

Phage-Mediated Selection on Microbiota of a Long-Lived Host

It is increasingly apparent that the dynamic microbial com- munities of long-lived hosts affect their phenotype, including resistance to disease [1–3]. The host microbiota will change over time due to immigration of new species [4, 5], interaction with the host immune system [6, 7], and selection by bacteriophage viruses (phages) [8], but the rela- tive roles of each process are unclear. Previous metage- nomic approaches confirm the presence of phages infecting host microbiota [8, 9], and experimental coevolution of bacteria and phage populations in the laboratory has demonstrated rapid reciprocal change over time [10, 11]. The key challenge is to determine whether phages influence host-associated bacterial communities in nature, in the face of other selection pressures. I use a tree-bacteria-phage system to measure reciprocal changes in phage infectivity and bacterial resistance within microbial communities of tree hosts over one season. An experimental time shift shows that bacterial isolates are most resistant to lytic phages from the prior month and least resistant to those from the future month, providing clear evidence for both phage-mediated selection on bacterial communities and bacterial-mediated selection on phage communities in nature. These reciprocal changes suggest that phages indeed play a key role in shaping the microbiota of their eukaryotic hosts.
Show more

5 Read more

Genomic and Proteomic Characterization of the Broad-Host-Range Salmonella Phage PVP-SE1: Creation of a New Phage Genus

Genomic and Proteomic Characterization of the Broad-Host-Range Salmonella Phage PVP-SE1: Creation of a New Phage Genus

We observed that PVP-SE1 is able to infect all mutants except the Rd1 and Rd2 mutants with roughly the same effi- ciency of plating. The fact that infections with the wild-type strain and Ra mutants result in turbid plaques but that infec- tions of most mutants defective in core polysaccharide assem- bly result in clear plaques suggests that the true receptor for this phage is the LPS inner core region (Fig. 5), as in the broad-host-range temperate phages P1 and P2 (47). This fact is also supported by the fact that PVP-SE1 infects both E. coli BL21 and K-12 (47). The inner core region is conserved in many enterobacteriaceae and similar in Salmonella Typhimu- rium and some E. coli strains. As a consequence, the ability of the phage to use it as a receptor may account for PVP-SE1’s broad host range and also its ambivalent nature in infecting E. coli (28, 61). Moreover, phage Felix 01 binds to receptors in the LPS outer core (it is unable to infect the Rb2 mutant), a less conserved region, which may explain the broader spectrum of phage PVP-SE1 than that of Felix 01 (31, 61). Unlike phage T7, which also binds to the inner core of LPS, the presence of the LPS O antigen does not prevent PVP-SE1 from reaching and binding the host LPS inner core and consequently from producing lysis (47, 55). This enables the phage to infect both rough and smooth bacteria. While PVP-SE1 cannot infect the Rd1 and Rd2 mutants, it is able to infect the deep rough (heptoseless) Re mutants. This suggests that the phage, like coliphage T4, may recognize more than one surface receptor: LPS and probably an outer membrane protein which might have become more easily accessible through the lack of LPS, thus increasing the lytic spectrum of PVP-SE1 (75).
Show more

9 Read more

Crystallographic Evidence for the Host-Guest Interaction of Metallamacromolecules

Crystallographic Evidence for the Host-Guest Interaction of Metallamacromolecules

study in solution is informative, X-ray crystallography provides the clearest structural evidence of the subtle intermolecular interactions. The several mechanisms including guest inclusion, guest exchange, and host–guest interactions are really tedious to follow up because of dynamic and flexible nature of metallahosts. The another challenge to this field is to engineer a system that produces significant changes in selectivity, taking into account the many factors or interactions that may be reinforcing each other or may compete. Thus the functionalization and application of host–guest chemistry of metallamacromolecules still in its fancy and have to be carried out in the coming years. We believe that this highlight review will increase awareness of metallamacromolecules and accelerate the development of host-guest functional materials.
Show more

10 Read more

Metaproteome analysis reveals that syntrophy, competition, and phage-host interaction shape microbial communities in biogas plants

Metaproteome analysis reveals that syntrophy, competition, and phage-host interaction shape microbial communities in biogas plants

Overall 0.4% ± 0.3% (minimum 0.11%, maximum 1.21%) of the identified spectra were associated with viral proteins (Fig. 2, Additional file 12). The highest virus abundance was observed for the thermophilic BGPs, i.e., BGP_05a and BGP_05b (Fig. 5 and Additional file 7: Table S5). In contrast to Fig. 2, Additional file 12, the calculation of the phage abundance in Fig. 5 and Additional file 7: Table S5 considers also phage metaproteins that were assigned automatically on root level, only (Additional file 9: Figure S2). The manual reannotation of this large group accounting 77% of all identified viral spectra was carried out using descriptions of metaproteins indicating typical viral functions. Furthermore, phage metagenome se- quences from BGPs [13] were added to the reference data- base. But the number of identified phage proteins did not increase (data not shown). A large portion of phage pro- teins was identified based on single peptides matching from conserved domains. In future experiments, the iden- tification of phage proteins has to be improved by better matching phage metagenomes.
Show more

17 Read more

Host-vector interaction in dengue: a simple mathematical model

Host-vector interaction in dengue: a simple mathematical model

system could maintain its resistance to an epidemic during adverse weather conditions causing an increase of the mosquito population. It is important to realize, that not only mosquito eradication but measures elevating (n/N) c are also important in controlling dengue. An increase of k and/or h and decrease of a and/or b will raise the magnitude of (n/N) c . As the inverse of h is the mean life time of infected mosquitoes, decreasing mosquito survival time will increase (n/N) c . The life time of Aedes aegypti is about two weeks to one month depending on environ- mental conditions. There is no evidence that acquiring the virus has a significant influence on longevity of Aedes aegypti. The effect of climatic conditions on survival of Aedes aegypti has been studied [21-22]. Extreme temperatures and torrential rains lower mosquito survival [23-24]. Nature of the microenvironment seems to play an important role in longevity of mosquitoes. It has been observed that longevity of Aedes aegypti is significantly higher in disorganized and thickly populated urban areas compared to more organized and planned urban dwellings [25]. Presence of moist shady resting and hiding places enhance mosquito breeding and survival. Consequently, according to the model, the critical mosquito population per person (n /N) c sufficient to initiate a dengue out-break in congested urban area is relatively low.
Show more

7 Read more

Structure, Adsorption to Host, and Infection Mechanism of Virulent Lactococcal Phage p2

Structure, Adsorption to Host, and Infection Mechanism of Virulent Lactococcal Phage p2

Taken together, our results illustrate that phage p2 infection of L. lactis MG1363 is likely a multistep process (Fig. 9). First, the tail tube adhesion domains may interact with host cell wall saccha- rides in order to maintain the phage close to its target, as proposed for phage ␭ or SPP1 (Fig. 9B). Once near its host, the external accessible RBP sites of the baseplate in the nonactivated confor- mation may scan the hosts surface (Fig. 9C) in search of specific pellicle phosphohexasaccharide motifs, which differ between dif- ferent L. lactis strains (61). If the number of specific binding events is large enough and Ca 2⫹ is available, the mechanical pull induced by this binding may destabilize the rest conformation of the base- plate and disrupt the interaction between the second Dit “arm” (14) and the RBP head domain. The RBPs would then initiate a 200° rotation toward the host’s surface, a movement coupled with Tal opening (14) (Fig. 9D). This large conformational change is probably somehow coupled to a signal to the stopper to open (26), releasing the dsDNA from the phage capsid. The dsDNA would in turn push out the TMP, which is suspected to guide delivery to the cytoplasm (54, 63, 64). In a majority of phages, the Tal protein harbors at its C terminus endolysin activity to digest the pepti- doglycan and form a hole permitting the passage of the MTP and dsDNA at the onset of infection. However, no peptidase sequence could be identified in the phage p2 genome as for the other phages of the 936 group (20). These phages infect their host only during the exponential phase, while those from the P335 subgroup II, having a peptidase, also infect the host in stationary phase (20, 65). The peptidoglycan is cross-linked during the stationary phase, and these P335 phages probably need the peptidase activity for their DNA to penetrate the cell wall.
Show more

11 Read more

Host-phage interaction on Agrobacterium tumefaciens. IV. Phage-directed protein synthesis.

Host-phage interaction on Agrobacterium tumefaciens. IV. Phage-directed protein synthesis.

Three classes of protein were detected: early proteins, class I, which include a protein capable of shutting off host protein synthesis; class II, proteins which are detected after 30 mi[r]

7 Read more

The effect of ribonuclease on phage-host interaction

The effect of ribonuclease on phage-host interaction

The most reasonable interpretation of these facts seems to be that in the presence of ribonuclease phage and bacteria do com- bine, and that during this phase the phage can be inac[r]

9 Read more

Visualization of Host-Polerovirus Interaction Topologies Using Protein Interaction Reporter Technology

Visualization of Host-Polerovirus Interaction Topologies Using Protein Interaction Reporter Technology

model defined in this study (Fig. 3B) supports epitope mapping data that shows the N terminus of the PLRV CP domain in some virus particles is partially surface accessible in virions purified from infected tissue (59). Our host-virus interaction models pre- dict both BiP and PsbQ2 interacting with this R-rich domain, while ACO3 is predicted to interact with the surface of the virion (Fig. 6 to 8). Alternatively, these data may represent interactions with virions at different stages of assembly that copurified with intact virions (59, 68) or even multimeric forms of the structural proteins (58) that could act in trans to regulate PLRV infection. It is known that viral proteins can be a part of many different host protein complexes that are dynamically regulated throughout the course of an infection (73, 74). Indeed, Western blot analyses showed evidence supporting several distinct multimers contain- ing PLRV CP/RTD after cross-linker application, including a CP dimer not previously observed in PLRV purified from infected tissue (Fig. 1C). In addition, we identified cross-links between the P1 viral replicase of PLRV to multiple sites within RuBisCO but no direct association of the structural proteins with either P1 or RuBisCO, indicating the presence of multiple viral-host protein complexes within our samples. Although we used known struc- tural information gathered from X-ray crystallography in combi- nation with the distance constraints defined by the identified PIR cross-links to generate models that best fit all the observed host- virus and virus-virus interactions, it is also possible that the iden- tified cross-links could represent single connections within several different protein complexes whose topological features change over time or in the presence of posttranslational modification (71). Due to rapid hydrolysis of the cross-linker, PIR application in cells or to infectious particles captures protein complexes dur- ing a single moment in time and thus represents a “snap-shot” of all the possible protein-protein interactions that can occur. PIR cross-linking studies using mutational analysis to target viral and host residues predicted by our models to be within the site of protein binding, as well as infection time course studies, would help to further pinpoint the most accurate models and the func- tions of these interactions.
Show more

15 Read more

The Self-Organization of Interaction Networks for Nature-Inspired Optimization

The Self-Organization of Interaction Networks for Nature-Inspired Optimization

B. Other Interaction Networks in Evolutionary Algorithms In this initial investigation of self-organizing interaction networks in an Evolutionary Algorithm, we have intentionally focused on the simple but important interactions associated with competition and survival. There are other interaction types that are highly relevant and worth studying such as interactions associated with multi-parent search operations. At a smaller scale, one could also consider the evolution of interaction networks between genes in individual population members. Such work could follow a more traditional path of self-adaptation to create advanced search operators or one could consider less explored territory such as indirect gene expression (e.g. via Gene Regulatory Networks).
Show more

12 Read more

Persistent Gammaherpesvirus Replication and Dynamic Interaction with the Host In Vivo

Persistent Gammaherpesvirus Replication and Dynamic Interaction with the Host In Vivo

Gammaherpesviruses establish life-long persistency inside the host and cause various diseases during their persistent infection. However, the systemic interaction between the virus and host in vivo has not been studied in individual hosts continuously, although such information can be crucial to control the persistent infection of the gammaherpesviruses. For the noninvasive and continuous monitoring of the interaction between gammaherpesvirus and the host, a recombinant murine gammaherpesvirus 68 (MHV-68, a gammaherpesvirus 68) was constructed to express a firefly luciferase gene driven by the viral M3 promoter (M3FL). Real-time monitoring of M3FL infection revealed novel sites of viral replication, such as salivary glands, as well as acute replication in the nose and the lung and progression to the spleen. Continuous monitoring of M3FL infection in individual mice demonstrated the various kinetics of transition to different organs and local clearance, rather than systemically synchronized clearance. Moreover, in vivo spontaneous reactivation of M3FL from latency was detected after the initial clearance of acute infection and can be induced upon treatment with either a proteasome inhibitor Velcade or an immunosuppressant cyclosporine A. Taken together, our results dem- onstrate that the in vivo replication and reactivation of gammaherpesvirus are dynamically controlled by the locally defined interaction between the virus and the host immune system and that bioluminescence imaging can be successfully used for the real-time monitoring of this dynamic interaction of MHV-68 with its host in vivo.
Show more

12 Read more

Development of a novel high throughput method for identifying phage-host pairs in an extreme environment

Development of a novel high throughput method for identifying phage-host pairs in an extreme environment

Previous studies have attempted to understand phage diversity by using conserved phage genes as markers. A number of these studies used Gene 20 as a diversity maker. Gene 20 (g20) is a homologue of the T4 phage portal protein and is conserved among T4- like myoviruses with hosts ranging from Proteobacteria to Cyanobacteria (Mann et al., 2005; Sullivan et al., 2008). It was initially identified as a potential marker because its evolution was likely to be restrained due to the geometric precision with which its protein product initiates capsid assembly (Hsiao & Black, 1978; Rao & Black, 2005). However, studies have demonstrated, using DNA fingerprinting and PCR with non-degenerate primers, the large extent of genetic variation across different environments over time which remains unrepresented in cultures (Frederickson et al., 2003). One of these studies sequenced g20 genes from 38 marine myophages isolated using Synechococcus and Prochlorococcus hosts. The majority of those sequences clustered into distinct groups representative of myophages specific to the host from which they were isolated from, but about 10% formed a separate cluster indicative of other potential primary hosts (Sullivan et al., 2008). While the study established some correlation between the host of isolation, g20 sequence and the host range, it also highlighted the relative frequency of exceptions. This suggested that phage portal proteins might not be good indicators of phage hosts or the optimal habitat of the phage itself (Sullivan et al., 2008).
Show more

283 Read more

Klebsiella Phage ΦK64-1 Encodes Multiple Depolymerases for Multiple Host Capsular Types

Klebsiella Phage ΦK64-1 Encodes Multiple Depolymerases for Multiple Host Capsular Types

mants harboring both the deletion plasmid (e.g., ΔS2-2::pGEM-T-Km) and pKD46-DHFR were selected on LB agar supplemented with kanamycin at 50 ␮ g/ml and trimethoprim at 75 ␮ g/ml at 30°C. The modified host was cultured in LB medium containing 1 mM arabinose, which induced expression of the recom- binase system. Log-phase cultures of the induced bacteria (optical density at 600 nm ⫽ 0.5) were coincubated with various titers of ⌽ K64-1 for 30 min at 30°C. Mixtures of phage and bacteria were inoculated into top agar and then overlaid on LB agar plates. After overnight incubation at 30°C, single plaques were picked from the plate with moderate number of plaques ( ⬃ 30) for PCR confirmation to detect the presence of phage-borne mutant DNA resulting from recombination (first infection). Because wild-type/mutant mixtures may exist in a single plaque when the first infection was carried out, ca. 1 to 3 mutant-positive plaques were selected for the second infection to obtain single plaques again. The recombinant phage were further coincubated with Klebsiella for 30 min at 37°C, followed by use of the agar overlay method for isolation of a pure phage (second infection). The single plaques were subjected to PCR to confirm the identity of the mutant phage-forming plaques. Subsequently, ⌽ K64-1 deletion mutants were isolated and validated by PCR using multiple pairs of primers (Table 5 and Fig. 7). TABLE 5 Primers used in this study
Show more

16 Read more

Interaction between the sbcC gene of Escherichia coli and the gam gene of phage lambda.

Interaction between the sbcC gene of Escherichia coli and the gam gene of phage lambda.

T h e acquisition of an sbcC mutation conferred upon these strains essentially wild-type rates of growth and genetic recombination and resistance to DNA damaging agents; [r]

5 Read more

Influenza in surviving cells : a study in host-virus interaction

Influenza in surviving cells : a study in host-virus interaction

Whether influenza viruses are titrated in the allantois of whole eggs or in surviving bits of membra ne­ on-shell, variation from host to host is always present. In whole eggs the effect is readily demonstrable, but difficult to investigate as one test only can be made on any one egg. In surviving bits the experimental approach would present no difficulty, were it not for the fact that under optimal conditions of maintenance the egg to egg variation is minimal and requires prohibitively large tests for its demonstration. We have noticed however that under conditions which were below optimal, not only did the sensitivity of the technique drop but variation between eggs increased greatly. Indeed, often this was the first sign of suboptimal conditions. Such a combin­ ation of variable host resistance with the possibility of doing repeated tests on the same material should allow the study of what makes one set of cells more susceptible to infection than another.
Show more

211 Read more

Training host pathogen protein–protein interaction predictors

Training host pathogen protein–protein interaction predictors

Detection of protein-protein interactions (PPIs) plays a vital role in molecular biology. Particularly, pathogenic infections are caused by interactions of host and pathogen proteins. It is important to identify host-pathogen interactions (HPIs) to discover new drugs to counter infectious diseases. Conventional wet lab PPI detection techniques have limitations in terms of cost and large-scale application. Hence, computational approaches are developed to predict PPIs. This study aims to develop machine learning models to predict inter-species PPIs with a special interest HPIs. Specifically, we focus on seeking answers to three questions that arise while developing an HPI predictor: 1) How should negative training examples be selected? 2) Does assigning sample weights to individual negative examples based on their similarity to positive examples improve generalization performance? and, 3) What should be the size of negative samples as compared to the positive samples during training and evaluation? We compare two available methods for negative sampling: random vs. de novo sampling and our experiments show that de novo sampling offers better accuracy. However, our experiments also show that generalization performance can be improved further by using a soft de novo approach that assigns sample weights to negative examples inversely proportional to their similarity to known positive examples during training. Based on our findings, we have also developed an HPI predictor called HOPITOR (Host-Pathogen Interaction Predictor) that can predict interactions between human and viral proteins. The HOPITOR web server can be accessed at the URL: http://faculty.pieas.edu.pk/fayyaz/software.html#HoPItor.
Show more

16 Read more

Show all 10000 documents...