TABLE 2.

Number of times each promoter sequence is represented in the libraries

Library Name	No. of sequences deleted or under-representeda	Library size/1,024b				Estimated diversitye at:
		Sublibrary	Library	Transfectionc	Minimumd	P0	P3
−35	6	9	7	586	7	520	6
−30	2	17	304	527	17	670	49
−25	2	142	357	791	142	426	142
−20	4	12	28	371	12	571	46
−15	22	82	68	98	68	417	5
−10	4	180	11	98	11	433	1
−5	2	10	11	15	10	496	2
+1	2	50	49	27	27	688	11
+5	3	797	80	1270	80	589	71

^aThe number of sequences expected to be underrepresented or deleted, due to the use of XbaI, XhoI, BssHII, and SstI during construction and runoff transcription of the libraries. 

^bThe observed number of clones or infective centers, divided by 1024, the size of an ideal library with five randomized positions. 

^cThe number of infectious centers obtained after transfection (7) was used as a minimal estimate of the library size in transfected cells. 

^dSmallest library size observed, up to and including the transfection step, equivalent to the minimum average number of times a promoter sequence is expected to be in the virus population at P0. In all cases, there is a 0.999 or better probability (by the Poisson distribution, 1 − e^λ, where λ = the smallest library size observed) that all possible sequences in the randomized region are present in the P0 population. Subsequent passages, at a MOI of ≤0.1, used a minimum of 2,000 PFU. 

^eThe diversity (8) of each population at P0, immediately after transfection, and passage 3 is estimated by analysis of the sequence of individual isolates (see Materials and Methods).