Limits...
Fr-TM-align: a new protein structural alignment method based on fragment alignments and the TM-score.

Pandit SB, Skolnick J - BMC Bioinformatics (2008)

Bottom Line: The assessment showed that the structural alignment quality from Fr-TM-align is better in comparison to both CE and TM-align.On average, the structural alignments generated using Fr-TM-align have a higher TM-score (~9%) and coverage (~7%) in comparison to those generated by TM-align.Fr-TM-align uses an exhaustive procedure to generate initial seed alignments.

View Article: PubMed Central - HTML - PubMed

Affiliation: Center for the Study of Systems Biology, School of Biology, Georgia Institute of Technology, Atlanta, USA. spandit3@mail.gatech.edu

ABSTRACT

Background: Protein tertiary structure comparisons are employed in various fields of contemporary structural biology. Most structure comparison methods involve generation of an initial seed alignment, which is extended and/or refined to provide the best structural superposition between a pair of protein structures as assessed by a structure comparison metric. One such metric, the TM-score, was recently introduced to provide a combined structure quality measure of the coordinate root mean square deviation between a pair of structures and coverage. Using the TM-score, the TM-align structure alignment algorithm was developed that was often found to have better accuracy and coverage than the most commonly used structural alignment programs; however, there were a number of situations when this was not true.

Results: To further improve structure alignment quality, the Fr-TM-align algorithm has been developed where aligned fragment pairs are used to generate the initial seed alignments that are then refined using dynamic programming to maximize the TM-score. For the assessment of the structural alignment quality from Fr-TM-align in comparison to other programs such as CE and TM-align, we examined various alignment quality assessment scores such as PSI and TM-score. The assessment showed that the structural alignment quality from Fr-TM-align is better in comparison to both CE and TM-align. On average, the structural alignments generated using Fr-TM-align have a higher TM-score (~9%) and coverage (~7%) in comparison to those generated by TM-align. Fr-TM-align uses an exhaustive procedure to generate initial seed alignments. Hence, the algorithm is computationally more expensive than TM-align.

Conclusion: Fr-TM-align, a new algorithm that employs fragment alignment and assembly provides better structural alignments in comparison to TM-align. The source code and executables of Fr-TM-align are freely downloadable at: http://cssb.biology.gatech.edu/skolnick/files/FrTMalign/.

Show MeSH
Histogram showing the fraction of protein pairs with improved/decreased or unchanged TM-score by Fr-TM-align with respect to the TM-score reported by TM-align. dTM is defined as (TM-score (Fr-TM-align) – TM-score (TM-align)).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2628391&req=5

Figure 4: Histogram showing the fraction of protein pairs with improved/decreased or unchanged TM-score by Fr-TM-align with respect to the TM-score reported by TM-align. dTM is defined as (TM-score (Fr-TM-align) – TM-score (TM-align)).

Mentions: We present the results below only for dataset 2. Using this criterion (dTM ≥ 0.05), Fr-TM-align shows improvements in TM-score for ~28% of protein pairs being compared. Of the 19,900 pairs, ~56% of protein pairs have a dTM between 0.0 and 0.05, and for ~8%, the TM-score remains unchanged. We have also assessed the correlation between dTM and the TM-score from TM-align, to see if Fr-TM-align results in improved TM-scores for only a range of TM-score as reported from TM-align. dTM is divided into 5 bins (dTM < -0.05; -0.05 ≤ dTM < 0.00; dTM = 0.0; 0.00 < dTM ≤ 0.05 and dTM > 0.05). For a bin size of 0.1 of TM-score as reported from TM-align, we calculated the fraction of proteins in the above mentioned dTM bins. Figure 4 shows the fraction of proteins in each dTM bin for the range of TM-scores. It is evident that Fr-TM-align clearly improves the TM-score over all the ranges of TM-score as reported from TM-align. It is interesting to note that significant improvement (dTM > 0.05) is observed for more protein pairs in the lower ranges of TM-score. It was previously reported that using TM-align, the average TM-score for a randomly related protein structures is 0.30 [34]. In our study, of the 40% of protein pairs in the range of TM-score (0.2–0.3), ~35% of pairs of protein pairs show an improvement of more than 0.05 in TM-score by Fr-TM-align. This suggests that many pairs, which previously were defined as randomly related by TM-align, now have a significant TM-score. Next, we evaluated the dependence of dTM on the length of the proteins being aligned. Figure 5 shows the histogram of the fraction of proteins in various dTM bins versus the length of the smaller of the two proteins. It is evident from Figure 5 that improvement in TM-score by Fr-TM-align is not strongly dependent on the lengths of the structurally aligned proteins.


Fr-TM-align: a new protein structural alignment method based on fragment alignments and the TM-score.

Pandit SB, Skolnick J - BMC Bioinformatics (2008)

Histogram showing the fraction of protein pairs with improved/decreased or unchanged TM-score by Fr-TM-align with respect to the TM-score reported by TM-align. dTM is defined as (TM-score (Fr-TM-align) – TM-score (TM-align)).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2628391&req=5

Figure 4: Histogram showing the fraction of protein pairs with improved/decreased or unchanged TM-score by Fr-TM-align with respect to the TM-score reported by TM-align. dTM is defined as (TM-score (Fr-TM-align) – TM-score (TM-align)).
Mentions: We present the results below only for dataset 2. Using this criterion (dTM ≥ 0.05), Fr-TM-align shows improvements in TM-score for ~28% of protein pairs being compared. Of the 19,900 pairs, ~56% of protein pairs have a dTM between 0.0 and 0.05, and for ~8%, the TM-score remains unchanged. We have also assessed the correlation between dTM and the TM-score from TM-align, to see if Fr-TM-align results in improved TM-scores for only a range of TM-score as reported from TM-align. dTM is divided into 5 bins (dTM < -0.05; -0.05 ≤ dTM < 0.00; dTM = 0.0; 0.00 < dTM ≤ 0.05 and dTM > 0.05). For a bin size of 0.1 of TM-score as reported from TM-align, we calculated the fraction of proteins in the above mentioned dTM bins. Figure 4 shows the fraction of proteins in each dTM bin for the range of TM-scores. It is evident that Fr-TM-align clearly improves the TM-score over all the ranges of TM-score as reported from TM-align. It is interesting to note that significant improvement (dTM > 0.05) is observed for more protein pairs in the lower ranges of TM-score. It was previously reported that using TM-align, the average TM-score for a randomly related protein structures is 0.30 [34]. In our study, of the 40% of protein pairs in the range of TM-score (0.2–0.3), ~35% of pairs of protein pairs show an improvement of more than 0.05 in TM-score by Fr-TM-align. This suggests that many pairs, which previously were defined as randomly related by TM-align, now have a significant TM-score. Next, we evaluated the dependence of dTM on the length of the proteins being aligned. Figure 5 shows the histogram of the fraction of proteins in various dTM bins versus the length of the smaller of the two proteins. It is evident from Figure 5 that improvement in TM-score by Fr-TM-align is not strongly dependent on the lengths of the structurally aligned proteins.

Bottom Line: The assessment showed that the structural alignment quality from Fr-TM-align is better in comparison to both CE and TM-align.On average, the structural alignments generated using Fr-TM-align have a higher TM-score (~9%) and coverage (~7%) in comparison to those generated by TM-align.Fr-TM-align uses an exhaustive procedure to generate initial seed alignments.

View Article: PubMed Central - HTML - PubMed

Affiliation: Center for the Study of Systems Biology, School of Biology, Georgia Institute of Technology, Atlanta, USA. spandit3@mail.gatech.edu

ABSTRACT

Background: Protein tertiary structure comparisons are employed in various fields of contemporary structural biology. Most structure comparison methods involve generation of an initial seed alignment, which is extended and/or refined to provide the best structural superposition between a pair of protein structures as assessed by a structure comparison metric. One such metric, the TM-score, was recently introduced to provide a combined structure quality measure of the coordinate root mean square deviation between a pair of structures and coverage. Using the TM-score, the TM-align structure alignment algorithm was developed that was often found to have better accuracy and coverage than the most commonly used structural alignment programs; however, there were a number of situations when this was not true.

Results: To further improve structure alignment quality, the Fr-TM-align algorithm has been developed where aligned fragment pairs are used to generate the initial seed alignments that are then refined using dynamic programming to maximize the TM-score. For the assessment of the structural alignment quality from Fr-TM-align in comparison to other programs such as CE and TM-align, we examined various alignment quality assessment scores such as PSI and TM-score. The assessment showed that the structural alignment quality from Fr-TM-align is better in comparison to both CE and TM-align. On average, the structural alignments generated using Fr-TM-align have a higher TM-score (~9%) and coverage (~7%) in comparison to those generated by TM-align. Fr-TM-align uses an exhaustive procedure to generate initial seed alignments. Hence, the algorithm is computationally more expensive than TM-align.

Conclusion: Fr-TM-align, a new algorithm that employs fragment alignment and assembly provides better structural alignments in comparison to TM-align. The source code and executables of Fr-TM-align are freely downloadable at: http://cssb.biology.gatech.edu/skolnick/files/FrTMalign/.

Show MeSH