Density Peak clustering of protein sequences associated to a Pfam clan reveals clear similarities and interesting differences with respect to manual family annotation

Abstract

Background The identification of protein families is of outstanding practical importance for in silico protein annotation and is at the basis of several bioinformatic resources. Pfam is possibly the most well known protein family database, built in many years of work by domain experts with extensive use of manual curation. This approach is generally very accurate, but it is quite time consuming and it may suffer from a bias generated from the hand-curation itself, which is often guided by the available experimental evidence. Results We introduce a procedure that aims to identify automatically putative protein families. The procedure is based on Density Peak Clustering and uses as input only local pairwise alignments between protein sequences. In the experiment we present here, we ran the algorithm on about 4000 full-length proteins with at least one domain classified by Pfam as belonging to the Pseudouridine synthase and Archaeosine transglycosylase (PUA) clan. We obtained 71 automatically-generated sequence clusters with at least 100 members. While our clusters were largely consistent with the Pfam classification, showing good overlap with either single or multi-domain Pfam family architectures, we also observed some inconsistencies. The latter were inspected using structural and sequence based evidence, which suggested that the automatic classification captured evolutionary signals reflecting non-trivial features of protein family architectures. Based on this analysis we identified a putative novel pre-PUA domain as well as alternative boundaries for a few PUA or PUA-associated families. As a first indication that our approach was unlikely to be clan-specific, we performed the same analysis on the P53 clan, obtaining comparable results. Conclusions The clustering procedure described in this work takes advantage of the information contained in a large set of pairwise alignments and successfully identifies a set of putative families and family architectures in an unsupervised manner. Comparison with the Pfam classification highlights significant overlap and points to interesting differences, suggesting that our new algorithm could have potential in applications related to automatic protein classification. Testing this hypothesis, however, will require further experiments on large and diverse sequence datasets.

DPCfam: Unsupervised protein family classification by Density Peak Clustering of large sequence datasets

“…The output of the method is a set of metaclusters . In [ 16 ] and in Results, we provide evidence that sequences belonging to the same metacluster are likely to share a core evolutionary module and thus metaclusters can often be assimilated to protein families. DPCfam output can be used ‘as is’ or, alternatively, metaclusters can be the basis from which to build profile-HMMs for more sensitive searches of the sequence space [ 19 ].…”

“…Clustering of local alignments means that DPCfam can, in principle, identify families representing individual evolutionary modules or domains. The same algorithm was already described in [ 16 ], where it was applied to two relatively small Pfam clans (4,083 and 2,022 sequences, respectively). In this work, we utilize DPCfam to cluster the entire UniRef50 database (v. 2017_07), containing about 23 million sequences that share no more than 50% sequence identity between each other.…”

“… (A) for each query sequence in the reference dataset (UniRef50), we collect all search sequence regions (SSRs) that align to it (BLAST with E-value < 0.1); (B) we use Density Peak Clustering (DPC) to group together SSRs that align to similar portions of the query; (C) primary clusters obtained in (B) for all queries are “metaclustered” (using, again, DPC) according to the number of SSRs they share; (D) finally, metaclusters that still share a significant number of SSRs undergo a further merging step. Material from: ‘Russo et al, Density Peak clustering of protein sequences associated to a Pfam clan reveals clear similarities and interesting differences with respect to manual family annotation, BMC Bioinformatics, published 2021, BioMed Central Ltd.’ [ 16 ]. …”

“…We choose μ 1 = 0.2, according to the rule of thumb in [17]. The sensitivity of the results to the value of μ 1 and of the other parameters was benchmarked and discussed in [16]. We then compute the δ i of the SSRs,…”

“…DPCfam is based on local sequence similarity searches followed by clustering using Density Peak Clustering. While we previously published a small-scale experiment that focused on family classification of proteins contained in only two Pfam clans/superfamilies [ 16 ], this is the first time that we discuss the use of DPCfam for wholesale classification of a large (about 23 million sequences) sequence dataset. Performing this clustering required developing a dedicated pipeline designed from scratch to ensure memory scaling.…”