Bmo Other 1 Oth1
Emelia Lute
Hi Nico, I am looking for the same, I think Asana is missing this feature. I have been using Asana and I can say how widely flexible it is (which I love) in terms of creating tasks and assign them across several projects, creating tags etc. However when you go to a view to customize/sort the way you would like to see your/everyone tasks to be displayed it does not offer us that extraordinary flexibility. I hope that this is improved in the near future. At the moment, I have to spend extra time (costs) sorting tasks for myself and everyone.
I agree entirely with this! This is a standard in many other programs and it is really disappointing Asana has not addressed this yet. It should be quite clear that sorting, for example, by priority AND due date would be very helpful.
When sorting in a project, have the ability to sort by multiple columns. For instance, I have sorted the project by Assignee, but I would also like it to sort by due date. Currently, you can only pick one or the other.
Asana is a really great programme that we are using, however, I have found that when trying to filter down a project by multiple fields (e.g. client type, Region, priority etc.) when wanting to extract data into a CVS, I can only select 1 filter at a time. I think it would be a really useful feature to be able to add multiple filters to a search so that the information extracted can be more specific.
We would like to be able to filter tasks by multi selecting custom field values. Our tasks have one of multiple categories assigned via a custom field. It would be very useful to filter for more than only one of them in the list view, as well as in the advanced search. So far we were not able to come up with a workaround, which makes organizing big projects more complicated than necessary.
Docking algorithms are developed to predict in which orientation two proteins are likely to bind under natural conditions. The currently used methods usually consist of a sampling step followed by a scoring step. We developed a weighted geometric correlation based on optimised atom specific weighting factors and combined them with our previously published amino acid specific scoring and with a comprehensive SVM-based scoring function.
The scoring with the atom specific weighting factors yields better results than the amino acid specific scoring. In combination with SVM-based scoring functions the percentage of complexes for which a near native structure can be predicted within the top 100 ranks increased from 14% with the geometric scoring to 54% with the combination of all scoring functions. Especially for the enzyme-inhibitor complexes the results of the ranking are excellent. For half of these complexes a near-native structure can be predicted within the first 10 proposed structures and for more than 86% of all enzyme-inhibitor complexes within the first 50 predicted structures.
We were able to develop a combination of different scoring schemes which considers a series of previously described and some new scoring criteria yielding a remarkable improvement of prediction quality.
Protein-protein interactions and complex formation play a central role in a broad range of biological processes, including hormone-receptor binding, protease inhibition, antibody-antigen interaction and signal transduction [1]. As structural genomics projects proceed, we are confronted with an increasing number of proteins with a characterised 3D structure but without a known function. To identify how two proteins are interacting will be particularly important for elucidating functions and designing inhibitors [2]. Although predicting around 50 percent false positive interactions [3], high throughput interaction discovery methods, such as the yeast two hybrid system, suggest thousands of protein-protein interactions and therefore also imply that a large fraction of all proteins interact with other proteins [4].
Since many biological interactions occur in transient complexes whose structures often cannot be determined experimentally, it is important to develop computational docking methods which can predict the structure of complexes with a proper accuracy [5].
Docking algorithms are developed to predict in which orientation two proteins are likely to bind under natural conditions. They can be split in a sampling step followed by a scoring step. A collection of putative structural complexes is generated by scanning the full conformational space in the first step, taking only geometric complementarity in consideration. Afterwards the putative complexes are ranked according to scoring functions based on chemical and additional aspects of geometrical complementarity.
Neuvirth et al. [8] reported different interface propensities for different atom-types, and there are specific atoms which play a crucial role in interactions, e.g. by their ability to participate in H-bonds. To take these phenomena into account, we also optimised atom specific weighting factors for the scoring of complex structures and evaluated their ability to identifiy near-native structures. Furthermore the combination of the amino acid specfic scoring with the atom specific scoring is evaluated.
As already described for the optimisation of the amino acid specific factors, we optimised the atom specific weighting factors for antibody-antigen, enzyme-inhibitor and 'other' complexes following the classification of the docking-benchmark2.0 [12]. Since the optimisation of a factor for each atom-type being present in proteins would have exceeded our computational resources we used the well established atom classification system by Melo et al. [13] consisting of 40 distinct atom types. The optimisation was accomplished using the nonlinear minimisation method (nlm) from the R-package for statistical computing [14].
In parallel to the development of the optimised weighting factors a very successfull comprehensive SVM-based scoring function was developed in our group and is described elsewhere (Martin O. and Schomburg D.; Efficient Comprehensive Scoring of Docked Protein Complexes using Probabilistic Support Vector Machines; submitted 2007) [15]. For this scoring function a support vector machine was trained to combine several scoring functions which were described to be able to identify near-native complexes (e.g. specialised energy functions, evolutionary relationship, class specific residue interface propensities, gap volume, buried surface area, empiric pair potentials on residue and atom level as well as measures for the tightness of fit). The application of the SVM-based scoring function leads to a remarkable improvement of the prediction quality as shown in table 2.
However, there is no factor included in this scoring function which directly describes the geometric fit of the two binding proteins. Thus we also show the results of the combination of the SVM-based scoring function with the scoring based on the weighted geometric correlation.
Optimised atom specific weighting factors. Colour coded atom specific weighted factors mapped on the 2D structures of the amino acids for enzyme-inhibitor and antibody-antigen complexes. The corresponding figure for 'other' complexes can be found within the supplementary material (additional file 2: supp_atm_factors_other.pdf). The numbers next to the atoms indicate the atom class.
For most of the other atom types the obtained parameters differ between the three complex classes. For antibody-antigen and for the 'other' complexes especially atoms that can be part of a hydrogen bond got higher values, while atoms that mainly contribute to the shape of the interface, like the methylene groups of the longer side-chains got low and very low values.
Especially for those atoms which are only present in the short side chains of ILE, LEU and VAL the optimisation yielded higher values for enzyme-inhibitor complexes than for the antibody-antigen complexes.
The values optimised for the interior of the receptor (I1) slightly differ for the three classes. For enzyme-inhibitor complexes the value is 0.67, for antibody-antigen complexes -0.87 and for the 'other' complexes 0.70.
The results of the application of the atom specific weighting factors is shown in table 1 and in figure 2. Figure 2 illustrates the strong enrichment of near native structures within the top 10% of the sorted prediction and in table 1 the number of complexes is shown for which a near native structure is found on the first rank and within the top10, top50, and top100 ranks. Furthermore the average rank for the first near-native structure (RMSD of the interface Cα atoms below 5 ) for each complex is given. Table 1 and figure 2 both show the improvement of prediction quality, compared to the results obtained from a purely geometric ranking, for the reranking based on the previously published amino acid specific weighting factors, for the atom specific weighting factors, and for a ranking based on the arithmetic mean of amino acid and atom specific scores.
The results achieved by the application of the atom specific weighting factors are comparable to those results which can be achieved with the amino acid specific scoring. Depending on the way of quality measurement the atom specific reranking is slightly better than the amino acid specific weighting. Especially the number of complexes for which a near native structure can be found on the top ranks (table 1) is higher with the atom specific factors and the average rank of the first near native structure is considerably lower.
For the four antibody-antigen and for the four 'other' complexes in the validation set the performance of the weighted scoring is worse than the purely geometric scoring with respect to the number of complexes with a near native prediction on the top ranks. However for both of them the average rank of the first near native structure is also considerably lower with the atom specific scoring than with the amino acid specific scoring.
The results obtained by a scoring with the mean of amino acid and atom specific scores are nearly identical to those which can be obtained by the atom specific scoring. Since the enrichment of near native structures on the lower ranks (figure 1) is slightly higher with the combination of both scoring schemes this combination is used for further scorings.