Sequence-based analyses of ARS A

Additional Proteins

The following proteins are additionally used for the prediction of transmembrand alpha-helices and signal peptides and for the prediction of GO Terms:

BACR

BACR_HALSA is a bacterial membrane protein...

type	Position '	Description
Topological domain	14 – 23	Extracellular
Transmembrane	24 – 42	Helical; Name=Helix A
Topological domain	43 – 56	Cytoplasmic
Transmembrane	57 – 75	Helical; Name=Helix B
Topological domain	76 – 91	Extracellular
Transmembrane	92 – 109	Helical; Name=Helix C
Topological domain	110 – 120	Cytoplasmic
Transmembrane	121 – 140	Helical; Name=Helix D
Topological domain	141 – 147	Extracellular
Transmembrane	148 – 167	Helical; Name=Helix E
Topological domain	168 – 185	Cytoplasmic
Transmembrane	186 – 204	Helical; Name=Helix F
Topological domain	205 – 216	Extracellular
Transmembrane	217 – 236	Helical; Name=Helix G
Topological domain	237 – 262	Cytoplasmic

RET 4

• RET4_HUMAN is a human retinal-binding protein. It delivers retionl from the liver stores to the peripheral tissues. Defects can cause night vision problems.

type	Position '	Description
Signal peptide	1 - 18
Chain	19 - 201	Retinol-binding protein 4
Chain	19 - 200	Plasma retinol-binding protein (1-182)
Chain	19 - 199	Plasma retinol-binding protein (1-181)
Chain	19 - 197	Plasma retinol-binding protein (1-179)
Chain	19 - 194	Plasma retinol-binding protein (1-176)

INSL 5

• INSL5_HUMAN

LAMP 1

• LAMP1_HUMAN

A 4

• A4_HUMAN

Secondary Structure Prediction

DSSP

• DSSP file of ARS A

Prediction of Disordered Regions

DISOPRED

• DISOPRED Server

POODLE

• POODLE Server

IUPRED

• IUPRED Server

Meta-Disorder

• PredictProtein Server

Prediction of transmembrane alpha-helices and signal peptides

SignalP

ARS A	A4	RET4	INSL5	LAMP1	BACR

TMHMM

TMHMM predicts transmembrane helices (TMH) using a Hidden Markov Model (HMM). The protein described by TMH model essentially consists of seven different states. Globular domains can occur on the cytoplasmic and the non-cytoplasmic side. On the cytoplsmic side, globular domains are linked to loops, ehich are agin linked to cytoplasimc caps. These caps are followed by the helex core and there is again a cap on the non-cytoplasmic side. These caps are linked to globular domains by either short or long non-cytoplasmic loops.
TMHMM outputs the most likely structure of the protein, ragarding to the above model. It also includes the orientation (cytoplasmic or non-cytoplasmic side) of the N-terminal signal sequence. The ouput consists of a plot - graphically showing the different states along the protein - and some additional statistics <ref> http://www.cbs.dtu.dk/services/TMHMM-2.0/TMHMM2.0.guide.html#output </ref>:

• The number of predicted transmembrane helices.
• The expected number of amino acids in transmembrane helices. If this number is larger than 18 it is very likely to be a transmembrane protein (OR have a signal peptide).
• The expected number of amino acids in transmembrane helices in the first 60 amino acids of the protein. If this number more than a few, you should be warned that a predicted transmembrane helix in the N-term could be a signal peptide.
• The total probability that the N-term is on the cytoplasmic side of the membrane.

Protein	#predicted TMHs	#expected AAs in TMHs	#expected AAs in TMHs in first 60 positions	orientation (N-term at non-cyto. side)	Graphical output
ARS A	0	2.65106	2.63079	0.12149
A4_HUMAN	1	22.72525	0.0027	0.00015
INSL5_HUMAN	0	0.50415	0.50415	0.03772
LAMP1_HUMAN	2	44.89582	22.24286	0.99287
RET4_HUMAN	0	0.01196	0.01179	0.01909
BACR	6	140.4032	26.1196	0.01887

Discussion

• ARS A:outside 1 507 (=all)

• A4_HUMAN: The topology is given below

Description	Position '
outside	1-700
TMhelix	701-723
inside	724-770

• INSL5_HUMAN: outside 1 135 (all residues)

• LAMP1_HUMAN POSSIBLE N-term signal sequence
  

Description	Position '
inside	1-10
TMhelix	11-33
outside	34-383
TMhelix	384-406
inside	407-417

• RET4_HUMAN: outside 1 201 (all)
• BACR:
  

1. sp_P02945_BACR_HALSA POSSIBLE N-term signal sequence

Description	Position '
outside	1-22
TMhelix	23-42
inside	43-54
TMhelix	55-77
outside	78-91
TMhelix	92-114
inside	115-120
TMhelix	121-143
outside	144-147
TMhelix	148-170
inside	171-189
TMhelix	190-212
outside	213-262

OCTOPUS and SPOCTOPUS

OCPTOPUS uses a combination of a Hidden Markov Model and neural network to predict the topology of a transmembrane protein. It uses BALST to create a sequence profile, whihc is then used by the neural network to predict the preference of the amino acids to be located within a transmembrane (M), interface (I), close loop (L) globular loop (G), inside (i) or outside (o). These scores are then passed to the HMM, which predicts the final states.
SPOCTOPUS extends the OCTOPUS algorithm with a preprocessing step. OCTOPUS does not predict signal peptides. The N-terminal targeting sequences mainly consist of hydrophobic residues and thus thier properties strongly resemble the transmembrane helices. Not considering the signal peptides in the prediction often leads to a false prediction of a transmembrane helix at the N-terminal domain. Therefore SPOCTOPUS extends the OCTOPUS algorithm with the prediction of signal peptide preference scores within the first 70 amino acids of the protein. The exact location of a potential signal peptide are then predicted by a HMM in OCTOPUS.

Protein	OCTOPUS	SPOCTOPUS
ARS A
A4
RET4
INSL5
LAMP1
BACR

TargetP

TargetP is used to predict the cellular localization of a protein. It combines the two methods ChloroP and SignalP. The following targeting sequences can be identified:

• chloroplast transit peptide (cTP)
• mitochondrial targeting peptide (mTP)
• secretory pathway signal peptide (SP)

TargetP uses a neural network to calculate and outputs scores for each of the above subcellular targets. TargetP finally predicts the location with the highest score. In our case all proteins are predicted to be targeted to the secretory pathway (S). Results are shown below. Note, that cTP is not included in our predictions, as we only considered eukaryotic and bacterial proteins. Also note, that TargetP is trained on eukaryotic proteins and hence the prediction for the protein "BACR", which is bacterial does not make sense, because there are completely different pathways of localization and secretion in eukayotes and bacteria (e.g. bacteria do not have an endoplasmatic reticulum, Golgi-Apparatus or Lysosome). Nevertheless, we included it in our analysis to see if TargetP predicts finds any localization sequence in it or predicts "-" (= no localization signal found).

Protein	mTP	SP	other	prediction
ARS A	0.070	0.926	0.054	S
A4_HUMAN	0.035	0.937	0.084	S
INSL5_HUMAN	0.074	0.899	0.037	S
LAMP1_HUMAN	0.043	0.953	0.017	S
RET4_HUMAN	0.242	0.928	0.020	S
BACR (bacterial)	0.019	0.897	0.562	S

Discussion

All proteins are assigned to the secretory pathway.

• Arylsulfatase A is a lysosomal enzyme. Therefore, the prediction is correct, as lysosomal proteins are guided there by the secretory pathway, via the endoplasmatic reticulum and the Golgi apparatus.
• coming
• coming
• coming
• coming
• As described above, BACR is a bacterial protein. TargetP assigns, that this protein is also guided to the secretory pathway, which makes no sense as the bacterial protein secretion is different from eukaryotic secretion. Nevertheless, the prediction is much less obvious in this case, compared to the others. The "other" class - meaning that no targeting sequence is found in the protein gets a considerable high score in this prediction, hence the assignment to S is more questionable here.

SignalP

sudo /apps/signalp-3.0/signalp -t gram- ../BACR.fasta > BACR.signalp
sudo /apps/signalp-3.0/signalp -t euk ../ARSA.fasta > ARSA.signalp
sudo /apps/signalp-3.0/signalp -t euk ../A4.fasta > A4.signalp
sudo /apps/signalp-3.0/signalp -t euk ../LAMP1.fasta > LAMP1.signalp
sudo /apps/signalp-3.0/signalp -t euk ../INSL5.fasta > INSL5.signalp
sudo /apps/signalp-3.0/signalp -t euk ../RET4.fasta > RET4.signalp

Prediction of GO Terms

GOPET

• GOPET Server

Pfam

• Pfam Server

ProtFun 2.2

• ProtFun 2.2 Server