The proteins obtained from BLASTP were analyzed for the presence of conserved domains using the PROSITE database (https://prosite.expasy.org/scanprosite/). According to this database, bacterial rhodopsins were specifically detected, and they showed two patterns: the first pattern illustrated the isoforms containing BACTERIAL_OPSIN_1, and the second pattern indicated BACTERIAL_OPSIN_RET. Two of the above-mentioned patterns were found in two isoforms of L. maculans (LMO-Q9HGT7 and LMSO-E4ZUL6), three isoforms of A. alternata (AAPOL-OWY42043.1, AAPOL-A0A177E1U0, and AAPOL-XP_018391147.1), S. sclerotiorum, and B. cinerea. The BACTERIAL_OPSIN_1 domain was found in one isoform of L. maculans (LMCAO-7BMH), one isoform of A. alternata (AAO1-A0A177E306), all isoforms of V. dahliae and V. longisporum, and nine isoforms of F. oxysporum (FOO1-RKK65588.1, FOCO1-KAG7002927.1, FOHP-RKK90771.1, FOHP-RKL21281.1, FOHPC-EGU75234.1, FOHPC-EGU78064.1, FOHPC-EGU79527.1, FOHPC-KAF6515179.1, and FOHPC-KAF6524808.1). Two isoforms of F. oxysporum contain the BACTERIAL_OPSIN_RET domain (FOHPF-QKD57451.1 and FOHPF-EWZ29335.1). Two other isoforms of F. oxysporum (FOHP-RKK62641.1 and FOHP-RKK62984.1) showed that Bac rhodopsin conserved domain in the Pfam database (http://pfam.xfam.org/) (Figure 1).

Several physicochemical and biochemical parameters of the opsin protein in L. maculans and six other fungal species were calculated using the ProtParam tool of the ExPASy. They are listed in Table 1. These parameters indicated the variation of the length of the opsin protein from 190 aa (VLHP-CRK17520.1) to 334 aa (FOHPF-EWZ29335.1). The lowest molecular weight was also found in V. longisporum species (VLHP-CRK17520.1), at 21166.72 kDa, and the highest one in F. oxysporum (FOHPF-EWZ29335.1), at 36625.13 kDa. The protein charge stability was investigated using isoelectric points (pI) (Gasteiger et al., 2005). The range of isoelectric points varies from 5.54 in B. cinerea to 9.08 in F. oxysporum. The instability index value was calculated for the experimental evaluation of a protein's stability. When the instability index values are lower than 40, the protein is predicted to be stable (Guruprasad et al., 1990). The calculation of the instability index revealed that all proteins were stable experimentally. The aliphatic index of a protein is the relative volume that is occupied by aliphatic side chains (alanine, valine, isoleucine, and leucine) (Ikai, 1980; Gasteiger et al., 2005). Three isoforms of A. alternata (AAPOL-OWY42043.1, AAPOL-A0A177E1U0, and AAPOL-XP_018391147.1) have a lower aliphatic index, whereas other isoforms have the highest thermostability, ranging from 100.81 to 125.47. To calculate the grand average of hydropathicity (GRAVY) of a peptide or protein, the sum of hydropathy values of all of the amino acids was divided by the number of residues in the sequence (Kyte and Doolittle, 1982; Gasteiger et al., 2005). The GRAVY analysis revealed that all isoforms of the opsin protein were hydrophobic, with a GRAVY value above 0.

The amino acid compositions of opsin proteins across species were compared using the heatmap package of the R program. This analysis shows that leucine, glycine, valine, isoleucine, threonine, and alanine have a greater portion of this protein than other amino acids (Figure 2). A comparison of the percentage amino acid composition of opsin proteins across all isoforms in terms of polar, nonpolar, positively and negatively charged amino acids showed that only one isoform of A. alternata, and all isoforms of S. sclerotiorum, B. cinerea, and L. maculans were enriched in nonpolar amino acids such as, alanine, valine and leucine, and polar amino acids such as, threonine. Three other isoforms of A. alternata have a high percentage of alanine, valine, and leucine from the non-polar amino acid group, along with aspartic acid from the negatively charged amino acid group.

The complex of two species of Verticillium (V. dahliae and V. longisporum) indicated the highest percentage of non-polar amino acids, including alanine, valine, leucine, glycine, and serine from the polar amino acid group. The percentage of amino acids in F. oxysporum is divided into three groups: the first group includes FOHPC-EGU79527.1, FOHPC-KAF6515179.1, FOHP-RKK90771.1, FOO1-RKK65588.1, FOHPC-KAF6524808.1, FOCO1-KAG7002927.1, FOHP-RKL21281.1, and FOHPC-EGU78064.1 and is rich in non-polar amino acids, including alanine, valine, isoleucine, glycine, and threonine from a polar group of amino acids. The second group of F. oxysporum (FOHPF-EWZ29335.1, FOHPF-QKD57451.1, FOHP-RKK62641.1, and FOHP-RKK62984.1) has a high percentage of non-polar amino acids, including alanine, valine, leucine, and isoleucine, and polar amino acids, including threonine and serine. The last isoform of F. oxysporum (FOHPC-EGU75234.1) presents a high percentage of alanine, valine, and leucine from non-polar amino acids, and threonine from polar amino acids, arginine as a positive amino acid, and aspartic acid as a negative amino acid (Supplementary Figure S1). The coevolutionary amino acid analysis revealed that tryptophan had been conserved among all species (Supplementary Figure S2).

Subcellular localization of the opsin protein of all isoforms was conducted using Loctree3 and illustrated that the majority of these proteins were localized in the endoplasmic reticulum membrane except for the two isoforms of V. dahliae (VDUP-A0A366PK01 and VDHP-RBQ92963.1) and five isoforms of V. longisporum (VLHP-CRJ88227.1, VLPLP-KAG7108306.1, VLPLP-KAG7111666.1, VLPLP-KAG7132832.1, and VLPLP-KAG7149394.1), which were localized in the mitochondrial membrane.

The other subcellular compartment is the nucleus. Proteins are synthesized in the cytoplasm, and they may be transported to the nucleus if they are needed there. There are binding sites on the protein sequence, known as nuclear-localized signals (NLSs), that mediate the protein's transportation to the nucleus (Brameier et al., 2007; Lu et al., 2021). Nuclear-localized signal analysis was conducted using the NucPred program. This analysis revealed that three isoforms of A. alternata (AAPOL-OWY42043.1, AAPOL-A0A177E1U0, and AAPOL-XP_018391147.1) and all isoforms of Verticillium, except for isoform VLHP-CRK17520.1, possessed NLS (Table 2). Transmembrane domain analysis was conducted using the PolyPhobius program. This analysis illustrates that most isoforms possess seven transmembrane domains, except for VLHP-CRK17520.1 with five TMs, FOHP-RKK62641.1, FOHP-RKK62984.1, and FOHP-RKL21281.1 that possess six TMs, and FOHPF-QKD57451.1 with eight TMs (Table 2; Supplementary Figure S3). This analysis shows that one sole isoform possesses a signal peptide (B. cinerea). Three isoforms of A. alternata (AAPOL-OWY42043.1, AAPOL-A0A177E1U0, and AAPOL-XP_018391147.1), S. sclerotiorum, and all isoforms of Verticillium, except for two isoforms, VDHP-KAF3356117.1 and VLHP-CRK17520.1, possess N-glycan. Four isoforms of F. oxysporum (FOHPF-QKD57451.1, FOHPF-EWZ29335.1, FOHP-RKL21281.1, and FOHPC-EGU78064.1) possess N-glycan, and FOHPC-EGU75234.1 shows two sites of N-glycan in its sequence.

An alignment file generated from multiple sequence alignment using the ClustalW tool in MEGAX, besides two outgroups, was used to construct a phylogenetic tree and analyze the evolutionary relationships of the opsin protein among different species (Figure 3). In the phylogenetic tree, 31 isoforms of the opsin protein from different fungal species were clustered into six major groups and six subgroups. Clades 1 and 2 included F. oxysporum and were divided into two subgroups. Two different species (S. sclerotiorum and B. cinerea) were grouped in the third clade. Clade 4 was divided into two subgroups: one subgroup consisted of one isoform of A. alternata (AAO1-A0A177E306), and three isoforms of L. maculans were grouped in the second subgroup of this clade. Clade 5 was also divided into two subgroups: the other isoforms of A. alternata were grouped in one subgroup of this clade, and one isoform of F. oxysporum (FOHPC-EGU75234.1) was in the other subgroup of this clade. Clade 6 consisted of a complex of two species of Verticillium: V. dahliae and V. longisporum.

A maximum parsimony tree was also created using the above-mentioned alignment file. The results showed identical main clades as the above-mentioned tree (Supplementary Figure S6).

The conserved motifs present in fungal species were determined by subjecting the amino acid sequence of the opsin protein to the MEME program. This analysis represented the fluctuation of motifs in the opsin protein during evolution. One highly conserved motif (motif 1) was found in all species. Further investigation also revealed that the conserved domain BACTERIAL_OPSIN_1 was located within motif 1. The second conserved domain, BACTERIAL_OPSIN_RET, which was found in two isoforms of L. maculans, three isoforms of A. alternata, S. sclerotiorum, and B. cinerea, was located within motif 4, whereas this domain of two isoforms of F. oxysporum was not detected within these motifs (Figure 4).

According to the conserved motifs and conserved domains analyses, two motifs (1 and 4) were shared in almost all isoforms, which illustrated the importance of these motifs in the opsin protein across all fungal species. Then, Weblogo 3 was used to construct the consensus sequences of these two motifs. Motif 1 was 50 aa long and was present across all species. Amino acid residue number 12 (tryptophan) was conserved in all species except for four species of F. oxysporum (FOHPF-QKD57451.1, FOHPF-EWZ29335.1, FOHP-RKK62641.1, and FOHP-RKK62984.1). In contrast, amino acid residue number 24 (cysteine) was more conserved in these four isoforms of F. oxysporum as well as two other isoforms of F. oxysporum (FOHP-RKL21281.1, FOHPC-EGU78064.1) and two species of Verticillium (V. dahliae and V. longisporum) (Supplementary Figure S4).

Motif 4 was 40 aa long and was present in all isoforms, except for the two isoforms of F. oxysporum (FOHP-RKK62641.1 and FOHP-RKK62984.1) (Supplementary Figure S5). The amino acid residue number 13 (tyrosine) was entirely conserved in all isoforms.

The prediction of the secondary structure features was performed using Jpred4 software and Proteus2 software. According to this prediction, the lowest and highest percentages of helix content were detected in FOHP-RKL21281.1 (35 %) and FOHPF-QKD57451.1 (74 %), respectively. The percentage of helix content of other isoforms ranged from 43 % to 64 %. Most varieties of beta-sheet content and coil content were observed in F. oxysporum: isoform FOHP-RKK62641.1 did not possess any beta-sheets, whereas the highest percentage of beta-sheet belonged to isoform FOHP-RKL21281.1 (18 %); the highest content of coil also belonged to isoform FOHP-RKL21281.1 (48 %), and isoform FOHPF-QKD57451.1 had the lowest content of coil (25 %) (Table 3).

The tertiary structure of all isoforms was predicted using SWISS-MODEL software. The conformational changes were observed in four isoforms of F. oxysporum (FOHPF-QKD57451.1, FOHPF-EWZ29335.1, FOHP-RKK62641.1, and FOHP-RKK62984.1) (Supplementary Figure S3).

The analysis of exon-intron structure led to insight into the evolution of this gene. Introns were not identified across all isoforms.

Biomolecular networks provide information regarding the molecular functions and interpretation of genomic variation. There are several biomolecular networks according to different purposes and scopes, including (1) networks of gene-gene regulatory events in transcription, networks of kinases/phosphatases, networks of metabolites and (2) networks of protein-protein interactions. Protein–protein interactions can occur under various physiological conditions and in different cell types. Proteins can also interact with each other indirectly, such as through a transcription factor that regulates gene expression and the production of another protein (Szklarczyk et al., 2019, 2021). The network of protein–protein interactions of the opsin protein of all isoforms was constructed using STRING software (version 11.5) (Figure 5).

Gene ontology (GO) is introduced as a common language in many organisms. GO is used to investigate and describe the function of proteins. The combination of gene ontology assessment and protein–protein interaction analysis enables the prediction of functions for unknown proteins. The three categories of GO are as follows: (1) biological process, which demonstrates the contribution of genes or proteins to specific biological events; (2) molecular function, which highlights the biochemical activities of the protein, such as ligand binding ligands, and (3) cellular component, which identifies the active location of a protein within cells. There is still a lack of knowledge regarding each category of GO for several proteins (Ashburner et al., 2000; Deng et al., 2004; Jain and Bader, 2010). An investigation into GO was conducted across all isoforms using STRING software (version 11.5) (Supplementary Table S1). Gene set enrichment analysis using STRING software showed no significant results.

The binding sites of the opsin protein were investigated using the P2Rank program. The total number of ligands in L. maculans ranged between 12 and 13, along with four pockets. Two ligands and six pockets were observed in three isoforms of A. alternata, and isoform AAO1-A0A177E306 of A. alternata possessed five ligands and six pockets. One identical ligand was observed in two species, S. sclerotiorum, and B. cinerea; they showed four and six pockets, respectively. V. dahliae and V. longisporum also possess a similar ligand with five pockets. The number of ligands and pockets in F. oxysporum ranged between 0 and 3 and 4 and 10, respectively (Table 3).

The initial step in fungal-plant interactions is the recognition of the appropriate plant host. Various signals (chemical or physical signals) are perceived by fungi and in turn, they respond by differentiation, movement to an appropriate infection site, and formation of invasion-related structures (Kumamoto, 2008; Bonfante and Genre, 2010). For example, Magnaporthe grisea, the causative agent of rice blast, responds to contact with an appropriate surface by producing structures named appressoria. GPCR which is encoded by PTH11 in this fungus, interacts with G proteins. Three subunits of G proteins (3 Gα subunits, 2 Gβ subunits, and one Gγ subunit) are encoded by the genome of M. grisea. And it has been reported that the formation of appressoria will be incomplete under the condition of deletion of each of these subunits (Nishimura et al., 2003; Kulkarni et al., 2005). In this research, opsin, a member of the rhodopsin family (the first family of GPCRs), of six phytopathogenic ascomycetous fungi (L. maculans, A. alternata, S. sclerotiorum, B. cinerea, V. dahliae, V. longisporum, and F. oxysporum) was studied.

Protein domains are evolutionarily conserved sequences with the basic components of protein structure and function (Murzin et al., 1995; Buljan et al., 2010). Evolutionary processes can lead to the emergence of new domains within proteins. Consequently, these new domains may contribute to the development of novel functions in organisms, potentially expanding their functional repertoire (Itoh et al., 2007). In comparison, fungi possess different numbers of domains than other organisms. Domain duplication during recombination causes the expansion of the domain in the fungal genome (Björklund et al., 2006). A study identified the bacteriorhodopsin domain in the microbial opsin homolog Sop1 in S. sclerotiorum (Lyu et al., 2016). Seven putative transmembrane helix domains have been reported in an opsin gene, nop-1, in N. crassa (Bieszke et al., 1999a). BACTERIAL_OPSIN_1 and BACTERIAL_OPSIN_RET are the two identified domains in this study. Most isoforms of studied fungi with the same domain/s were grouped in the same clade in the phylogenetic tree, which is hypothesized to have an identical biological function. Further analysis revealed that BACTERIAL_OPSIN_1 was located within the conserved motif of all species, indicating the importance of this conserved domain during evolutionary events. Bacterial opsins play a role in Light-dependent ion transport and sensory in a family of halophilic bacteria. These proteins are integral membrane proteins that contain seven transmembrane domains. The two classes of these proteins are as follows: (1) light-driven proton pumps, bacteriorhodopsin and archaerhodopsin and (2) a light-driven chloride pump, halorhodopsin (Oesterhelt, 1989; Soppa et al., 1993; Purschwitz et al., 2006).

The membrane composition influences the membrane protein's function and stability (Lee, 2004; Hunte and Richers, 2008; Phillips et al., 2009). Tryptophan is an amino acid that highly affects protein folding (Sanchez et al., 2008; Fiedler et al., 2010). Tryptophan is also involved in anchoring the hydrophobic transmembrane part across the membrane and causing hydrophobic mismatch, which highly affects the change of the transmembrane segment and the membrane structure (Sun et al., 2008). Transmembrane proteins also possess tryptophan amino acid near the end of the transmembrane helices. The above-mentioned functions of tryptophan are also reported in transmembrane proteins. Therefore, it is presumed that this amino acid could influence the function of opsin proteins as one of the transmembrane proteins (Von Heijne, 2007; De Jesus and Allen, 2013). Our study revealed that tryptophan was highly conserved across all isoforms (Supplementary Figure S2). Consensus sequence analysis of motif 1 demonstrated that tryptophan amino acid (residue 12) was highly conserved among all clades of the phylogenetic tree except for clade 2 (Supplementary Figure S4). According to the secondary and tertiary structure analysis of isoforms in clade 2 (FOHPF-QKD57451.1, FOHPF-EWZ29335.1, FOHP-RKK62641.1, and FOHP-RKK62984.1), the conformation of these isoforms has changed, confirming the importance of tryptophan regarding its effect on the structure of the protein membrane. Hence, it is hypothesized that these isoforms also have different functions.

L. maculans belongs to the Dothideomycetes class and the Pleosporales order. Several other important pathogens are in this order, such as those from the genus Alternaria. Pseudothecia production has been observed among most fungi of the Dothideomycetes class, forming either a single or a group of fruiting bodies. Fungi in the Dothideomycetes class have varieties of host plants, but their life cycles and pathogenicity strategies are often similar (Eyal, 1999; Nolan et al., 1999; Kaczmarek and Jedryczka, 2012). L. maculans is introduced as a hemibiotrophic fungus. Several plants were suggested as hosts for L. maculans in the past, such as Phaseolus sp., Humulus sp., Swertia perennis, Teucrium sp., and Artemisia campestris. Afterward, reports highlighted crucifers and mainly Brassica crops as the hosts of L. maculans (Mendes-Pereira et al., 2003). Fungi belonging to the genus Leptosphaeria can cause allergic responses of types I and II in humans (Hasnain, 1993). The genus Alternaria from the Dothideomycetes class mainly includes saprophytic fungi; some species are plant pathogens. There is a broad host range, such as cereals, ornamentals, oil crops, vegetables (cauliflower, broccoli, carrot, potato, tomato), citrus, and apple. Several species of Alternaria cause cancer in mammals. Moreover, spores of this genus are also introduced as airborne allergens (Meena et al., 2010). According to the similarity among fungi of the Dothideomycetes class, it has been hypothesized that isoforms AAO1-A0A177E306 and isoforms of L. maculans, which are grouped in one clade in the phylogenetic tree, can play a similar role in sporulation and have the identical growth pathway. They could also be considered allergenic pathogens for humans. Although the other isoforms of A. alternata may have different sporulation pathways, they may also have a less allergic effect on humans.

Two necrotrophs and soilborne fungi, S. sclerotiorum (white mold fungus) and B. cinerea (gray mold fungus) are very close taxonomically. Their host range is broader than that of most other plant pathogens (Bolton et al., 2006; Williamson et al., 2007). Both fungi produce sclerotia that survive in the soil. They also produce melanized sclerotium, which is involved in their lifecycle by germinating vegetatively or sexually. Botrytis spp. also belongs to the Sclerotiniaceae family. Proteins encoded by the genomes of B. cinerea and S. sclerotiorum are 83% identical, which points to a very close relationship between these two genera (Groves and Loveland, 1953; Nees, 1973; Amselem et al., 2011). Despite many similarities in developmental and physiological features between these two fungi, they vary in their regulation and potential for sporulation. S. sclerotiorum produces ascospores and is dispersed via airborne spores, whereas B. cinerea produces ascospores, but its dispersal is predominantly via conidia. Furthermore, they are different in the regulation of sexual sporulation, with S. sclerotiorum being homothallic while B. cinerea is heterothallic (Bolton et al., 2006; Williamson et al., 2007). Most similarities between these two fungi confirm the phylogenetic analysis of this study, and the above-mentioned differences that have been found in this study suggest that the opsin protein may be involved in the melanized sclerotium pathway and can have an identical function in their growth and sporulation.

The widespread distribution of Fusarium species has been reported on plants, in soil, and in water as parasites, endophytes, or saprophytes. Fusarium causes several varieties of diseases in on-field, horticultural, ornamental, and forest crops, such as wilts, blights, rots, and cankers. Varieties of toxic secondary metabolites are produced by fusaria, such as trichothecenes and fumonisins, that are involved in the contamination of agricultural products; trichothecenes can also be a virulence factor in plant diseases. A few Fusarium species have been introduced as opportunistic human pathogens that cause corneal infections (O'Donnell et al., 2004). Life cycle, specialization, host adaptation, and specificity differ among Fusarium species. F. oxysporum has a broad host range, including monocotyledonous and dicotyledonous plants. Moreover, it can also infect immunocompromised patients and other mammals (O'Donnell et al., 2004; Ortoneda et al., 2004). F. oxysporum is an anamorphic species. It is a heterogeneous and polytypic morphospecies that is introduced as the most abundant soilborne fungus (Waalwijk et al., 1996; O'Donnell and Cigelnik, 1997; Snyder and Hansen, 2013). Most of the F. oxysporum isoforms were grouped in the same clade in the phylogenetic analysis, which displays similarity within this forma. Despite these similarities, several subgroups have been observed in this phylogenetic tree analysis of the opsin protein, which points to the different performances of the opsin protein within this species, demonstrating different recognition and function strategies of this protein on different hosts. It has been hypothesized that one isoform of F. oxysporum (FOHPC-EGU75234.1), which was grouped with three isoforms of A. alternata, can not be an allergenic pathogen to humans.

The genus Verticillium involves soilborne plant pathogenic species that cause vascular wilt diseases on a wide range of plants, especially in the temperate and subtropical regions of the world (Fradin and Thomma, 2006). The Verticillium genus comprises two plant pathogenic species: V. dahliae, which forms microsclerotia. V. albo-atrum, which produces dark resting mycelium, and the other four species are saprophytes and non-pathogenic species (V. nigrescens, V. nubilum, V. tricorpus, and V. theobromae) (Barbara and Clewes, 2003). V. dahliae causes verticillium wilt in a wide range of host plants, including woody, dicotyledonous species of herbaceous annuals and perennials (Bhat and Subbarao, 1999). Subsequent studies have revealed the occurrence of parasexual hybridization between V. albo-atrum and V. dahliae to evolve V. longisporum (Karapapa et al., 1997; Zeise and Von Tiedemann, 2001). Several studies have illustrated the host specificity of V. longisporum in brassicaceous hosts, whereas V. dahliae is restricted to non-brassicaceous hosts (Zeise and Von Tiedemann, 2001, 2002a,b). The phylogenetic analysis of this study separated Verticillium species from other fungi, demonstrating higher similarity between these two species. These two species were separated within their clade in this phylogenetic analysis of the opsin protein, suggesting that this protein acts differently among these two species of the genus Verticillium concerning plant–pathogen interaction, pathogenicity, and sporulation.

Protein sequences of opsins from L. maculans, A. alternata, S. sclerotiorum, B. cinerea, V. dahliae, V. longisporum, and F. oxysporum were retrieved from the NCBI and UniProt databases. Their gene IDs are listed in Supplementary Table S2. Then, BLASTP was conducted across the proteomes of the above-mentioned fungi using the opsin protein sequences. The identified opsin proteins were analyzed for the presence of conserved domains using the Pfam [a protein families database that identifies proteins by using two alignments and a profile hidden Markov model (HMM)] (Finn et al., 2014), SMART (a database that is used to identify and annotate the genetically mobile domains and analysis of their architectures) (Schultz et al., 2000), ENSEMBLE (a comprehensive database that provides information on the annotation of individual genomes of many species such as plants and microorganisms) (Birney et al., 2004), Conserved Domain Database (CDD) (the annotation of protein sequences is provided by this resource along with the location of conserved domain footprints) (Marchler-Bauer et al., 2011), and PROSITE (two signatures (patterns and generalized profiles) are used in this database to identify conserved regions) (Hulo et al., 2006) databases. Therefore, these databases provided comprehensively conserved domain analyses by applying different alignment and annotation methods. The position and scale of protein domains were drawn using Domain Graph (DOG) software (http://dog.biocuckoo.org/).

Physicochemical properties, including isoelectric point (pI), molecular weight (Mw), instability index (II), grand average of hydropathicity (GRAVY), and aliphatic index (AI), were investigated for identified opsin proteins using the ProtParam tool available on the ExPASy website (https://web.expasy.org/protparam/) and EMBOSS Pepstats (https://www.ebi.ac.uk/Tools/seqstats/emboss_pepstats/). The coevolution of amino acids was investigated using MISTIC (Mutual Information Server to Infer Coevolution) (http://mistic.leloir.org.ar/index.php).

Subcellular localization of opsin proteins was predicted using Loctree3 (https://rostlab.org/services/loctree3/). Nuclear localization signal prediction was conducted using NucPred (https://nucpred.bioinfo.se/cgi-bin/single.cgi). The integral transmembrane (TM) domains were investigated using PolyPhobius (https://phobius.sbc.su.se/poly.html) and Protter (http://wlab.ethz.ch/protter/start/). The prediction of signal peptide cleavage sites was performed using TargetP-2.0 (https://services.healthtech.dtu.dk/services/TargetP-2.0/).

Multiple sequence alignment of the opsin sequences of these fungal species was performed using the ClustalW tool in MEGAX. Then, the alignment file was used to create a phylogenetic tree alongside with Maximum Likelihood method (1000 repeats) with the parameters of the Jones-Taylor-Thornton (JTT) matrix-based model, uniform rates among sites, and partial deletion of gaps in MEGAX software. Opsin sequences of one species of yeast (Cryptococcus neoformans) and one species of archaea (Haloarcula sinaiiensis) were downloaded from the UniProt database to be used as outgroup in the phylogenetic tree. The alignment file was also used to generate a phylogenetic tree, alongside with Maximum parsimony method in MEGAX. In the second phylogenetic tree the calculation of bootstrap values was also at 1000 iterations with the parameters of the Subtree-Pruning-Regrafting (SPR) method and partial deletion of gaps in MEGAX software. Two above-mentioned outgroups were also used in the second phylogenetic tree.

Conserved motifs of opsin proteins across all fungal species were predicted using the MEME (https://meme-suite.org/meme/) tool with the maximum number of eight motifs, and the other parameters were set to default. The sequence logos of conserved motifs were generated using Weblogo 3 (http://weblogo.threeplusone.com/).

The features of the secondary structure were predicted using Jpred4 software (https://www.compbio.dundee.ac.uk/jpred/) and Proteus2 software (http://www.proteus2.ca/proteus2/), and the SWISS-MODEL software (https://swissmodel.expasy.org/) was used for the prediction of the fatures of the tertiary structure of opsin proteins in all fungal species. To investigate the exon and intron structures, blastx was performed at NCBI for each isoform.

Protein–protein interactions (PPI) of the opsin protein in all fungal species were constructed using String software (https://string-db.org/). Afterward, constructed networks were investigated considering the GO annotation, including Biological Processes (BP), Molecular Functions (MF), and Cellular Component (CC). String software was also used for Gene set enrichment analysis across all isoforms. Binding sites, including ligand-binding pockets of the opsin protein, were predicted using P2Rank (https://prankweb.cz/).