# -----------------------------------------------------
# Summary of all used filtering rules:
# -----------------------------------------------------



1) case insensitive patterns (perl regular expressions)
# -----------------------------------------------------

	# name looks like an aminoacid
	^(ala(nine)?|cys(teine)?|asp(aragine)?|glu(tamine)?|phe(nylalanine)?|gly(cine)?|his(tidine)?|ile|isoleucine|lys(ine)?|leu(cine)?|met(hionine)?|asn|aspartic acid|aspartate|pro(line)?|gln|glutamic acid|glutamate|arg(inine)?|ser(ine)?|thr(eonine)?|val(ine)?|trp|tryptophane|tyr(osine)?)(-| )?\(?\d+\)?$

	# type/class/greek-letter not enough for a full name
	^(type|beta|alpha|gamma|eps[iy]lon|zeta|kappa|class)( |-)(\d+|[ivx]+)$

	# kDa pattern
	^[\d.]+[ \-]?k(Da?|b(yte)?)$

	# name contains punktuation characters
	(^| )[.,!?:;]( |$)

	# name contains too many dots
	\.[^.]+\.[^.]+\.

	# name contains 'cell-type' key-words at the end
	(^| |-)(cells?|\w+cytes?)$

	# name contains number plus plural word
	^([\d.,]+|two|three|four|fife|six|seven|eight|nine|ten|eleven|twelve) \w+[^s]s$

	# name contains 'cell' at the beginning plus only one following word (like cell type or cell division)
	^(cells?)( |-)\w+$

	# word matches no letters
	^[^a-z]+$

	# too many consecutive digits
	\d{5}

	# name contains rare characters (#=;%°@$?!)
	[=%;#@$?!]|°.?[CF]

	# name seems to be a physical unit
	(\/|^\d+ ?)(cm2|[nm]?mol|ml|kg|l|h|mg|g|min)( |$)

	# name seems to be an author's name
	\bet( \.)? al\b



2) case sensitive patterns (perl regular expressions)
# -----------------------------------------------------

	# mutations
	^([ACDEFGHIKLMNPQRSTVWY]|Ala(nine)?|Cys(teine)?|Asp(aragine)?|Glu(tamine)?|Phe(nylalanine)?|Gly(cine)?|His(tidine)?|Ile|Isoleucine|Lys(ine)?|Leu(cine)?|Met(hionine)?|Asn|Aspartic acid|Aspartate|Pro(line)?|Gln|Glutamic acid|Glutamate|Arg(inine)?|Ser(ine)?|Thr(eonine)?|Val(ine)?|Trp|Tryptophane|Tyr(osine)?)\d{2,4}([ACDEFGHIKLMNPQRSTVWY]|Ala(nine)?|Cys(teine)?|Asp(aragine)?|Glu(tamine)?|Phe(nylalanine)?|Gly(cine)?|His(tidine)?|Ile|Isoleucine|Lys(ine)?|Leu(cine)?|Met(hionine)?|Asn|Aspartic acid|Aspartate|Pro(line)?|Gln|Glutamic acid|Glutamate|Arg(inine)?|Ser(ine)?|Thr(eonine)?|Val(ine)?|Trp|Tryptophane|Tyr(osine)?)$

	# aminoacids
	^(Ala|Cys|Asp|Glu|Phe|Gly|His|Ile|Lys|Leu|Met|Asn|Pro|Gln|Arg|Ser|Thr|Val|Trp|Tyr)$

	# aminoacids plus position
	^[ACDEFGHIKLMNPQRSTVWY]\d{3}$

	# name seems to be an internet address
	(^(ht|f)tp://)|(\.html?)|(\.(com|de|uk|il|it|gov|edu|es|fr|nl|dk|ru|jp|ca|net)$)

	# name looks like a persons name
	^[A-Z]\. ?[A-Z][a-z]{4,}$

	# name looks like a dna-sequence
	^[AGTC]{5,}$

	# name looks like an rna-sequence
	^[AGUC]{5,}$

	# name looks like an rna/dna-type
	^([a-z]+[RD]NAs?|RNAi)$

	# name seems to be a physical unit (2)
	^(pH|pK[asb]?|pI)( =? ?[\d.]+)?$
	
	# name seems to be nucleotide
	^(nt[- ]?\d+|d?[AGCTU][TDM]P|N[TDM]Ps?)$
	name looks like a nucleotide



3) complex rules
# -----------------------------------------------------

	# the last four letters of the sample centre are in the 'chemical compound endings database'

	# sample centre preceeds 'cell' or 'cells'

	# sample centre suceeds 'virus' or 'virus type \d+'
	
	# sample centre is only one token long and in common dictionary
	
	# sample centre is exactly two tokens long where first word is a number => cannot to end in the following pattern:
	(protein|gene|factor|hormone|homolog(ue)?|collagen|pump|antibody|precursor|molecule|isoform|receptor|regulator|proteasome|allele|peptide|keratin|cytokine|chemokine|activator|transporter|ribozyme|antigen|translocator|subunit|repressor|receptor|channel|inhibitor|enzyme|ase|in)$

	# sample centre preceeds '=' (seems to be part of equation)

	# sample centre has number as last token and is followed by a plural noun